title
stringlengths 7
246
| abstract
stringlengths 6
3.31k
|
|---|---|
Wav2Seq: Pre-training Speech-to-Text Encoder-Decoder Models Using Pseudo Languages | We introduce Wav2Seq, the first self-supervised approach to pre-train both
parts of encoder-decoder models for speech data. We induce a pseudo language as
a compact discrete representation, and formulate a self-supervised pseudo
speech recognition task -- transcribing audio inputs into pseudo subword
sequences. This process stands on its own, or can be applied as low-cost
second-stage pre-training. We experiment with automatic speech recognition
(ASR), spoken named entity recognition, and speech-to-text translation. We set
new state-of-the-art results for end-to-end spoken named entity recognition,
and show consistent improvements on 20 language pairs for speech-to-text
translation, even when competing methods use additional text data for training.
Finally, on ASR, our approach enables encoder-decoder methods to benefit from
pre-training for all parts of the network, and shows comparable performance to
highly optimized recent methods. |
OPT: Open Pre-trained Transformer Language Models | Large language models, which are often trained for hundreds of thousands of
compute days, have shown remarkable capabilities for zero- and few-shot
learning. Given their computational cost, these models are difficult to
replicate without significant capital. For the few that are available through
APIs, no access is granted to the full model weights, making them difficult to
study. We present Open Pre-trained Transformers (OPT), a suite of decoder-only
pre-trained transformers ranging from 125M to 175B parameters, which we aim to
fully and responsibly share with interested researchers. We show that OPT-175B
is comparable to GPT-3, while requiring only 1/7th the carbon footprint to
develop. We are also releasing our logbook detailing the infrastructure
challenges we faced, along with code for experimenting with all of the released
models. |
A Change Dynamic Model for the Online Detection of Gradual Change | Changes in the statistical properties of a stochastic process are typically
assumed to occur via change-points, which demark instantaneous moments of
complete and total change in process behavior. In cases where these transitions
occur gradually, this assumption can result in a reduced ability to properly
identify and respond to process change. With this observation in mind, we
introduce a novel change-dynamic model for the online detection of gradual
change in a Bayesian framework, in which change-points are used within a
hierarchical model to indicate moments of gradual change onset or termination.
We apply this model to synthetic data and EEG readings drawn during epileptic
seizure, where we find our change-dynamic model can enable faster and more
accurate identification of gradual change than traditional change-point models
allow. |
RANG: A Residual-based Adaptive Node Generation Method for Physics-Informed Neural Networks | Learning solutions of partial differential equations (PDEs) with
Physics-Informed Neural Networks (PINNs) is an attractive alternative approach
to traditional solvers due to its flexibility and ease of incorporating
observed data. Despite the success of PINNs in accurately solving a wide
variety of PDEs, the method still requires improvements in terms of
computational efficiency. One possible improvement idea is to optimize the
generation of training point sets. Residual-based adaptive sampling and
quasi-uniform sampling approaches have been each applied to improve the
training effects of PINNs, respectively. To benefit from both methods, we
propose the Residual-based Adaptive Node Generation (RANG) approach for
efficient training of PINNs, which is based on a variable density nodal
distribution method for RBF-FD. The method is also enhanced by a memory
mechanism to further improve training stability. We conduct experiments on
three linear PDEs and three nonlinear PDEs with various node generation
methods, through which the accuracy and efficiency of the proposed method
compared to the predominant uniform sampling approach is verified numerically. |
A Survey on Uncertainty Toolkits for Deep Learning | The success of deep learning (DL) fostered the creation of unifying
frameworks such as tensorflow or pytorch as much as it was driven by their
creation in return. Having common building blocks facilitates the exchange of,
e.g., models or concepts and makes developments easier replicable. Nonetheless,
robust and reliable evaluation and assessment of DL models has often proven
challenging. This is at odds with their increasing safety relevance, which
recently culminated in the field of "trustworthy ML". We believe that, among
others, further unification of evaluation and safeguarding methodologies in
terms of toolkits, i.e., small and specialized framework derivatives, might
positively impact problems of trustworthiness as well as reproducibility. To
this end, we present the first survey on toolkits for uncertainty estimation
(UE) in DL, as UE forms a cornerstone in assessing model reliability. We
investigate 11 toolkits with respect to modeling and evaluation capabilities,
providing an in-depth comparison for the three most promising ones, namely
Pyro, Tensorflow Probability, and Uncertainty Quantification 360. While the
first two provide a large degree of flexibility and seamless integration into
their respective framework, the last one has the larger methodological scope. |
A walk through of time series analysis on quantum computers | Because of the rotational components on quantum circuits, some quantum neural
networks based on variational circuits can be considered equivalent to the
classical Fourier networks. In addition, they can be used to predict the
Fourier coefficients of continuous functions. Time series data indicates a
state of a variable in time. Since some time series data can be also considered
as continuous functions, we can expect quantum machine learning models to do
many data analysis tasks successfully on time series data. Therefore, it is
important to investigate new quantum logics for temporal data processing and
analyze intrinsic relationships of data on quantum computers.
In this paper, we go through the quantum analogues of classical data
preprocessing and forecasting with ARIMA models by using simple quantum
operators requiring a few number of quantum gates. Then we discuss future
directions and some of the tools/algorithms that can be used for temporal data
analysis on quantum computers. |
A Sharp Memory-Regret Trade-Off for Multi-Pass Streaming Bandits | The stochastic $K$-armed bandit problem has been studied extensively due to
its applications in various domains ranging from online advertising to clinical
trials. In practice however, the number of arms can be very large resulting in
large memory requirements for simultaneously processing them. In this paper we
consider a streaming setting where the arms are presented in a stream and the
algorithm uses limited memory to process these arms. Here, the goal is not only
to minimize regret, but also to do so in minimal memory. Previous algorithms
for this problem operate in one of the two settings: they either use
$\Omega(\log \log T)$ passes over the stream (Rathod, 2021; Chaudhuri and
Kalyanakrishnan, 2020; Liau et al., 2018), or just a single pass (Maiti et al.,
2021).
In this paper we study the trade-off between memory and regret when $B$
passes over the stream are allowed, for any $B \geq 1$, and establish tight
regret upper and lower bounds for any $B$-pass algorithm. Our results uncover a
surprising *sharp transition phenomenon*: $O(1)$ memory is sufficient to
achieve $\widetilde\Theta\Big(T^{\frac{1}{2} + \frac{1}{2^{B+2}-2}}\Big)$
regret in $B$ passes, and increasing the memory to any quantity that is $o(K)$
has almost no impact on further reducing this regret, unless we use $\Omega(K)$
memory. Our main technical contribution is our lower bound which requires the
use of information-theoretic techniques as well as ideas from round elimination
to show that the *residual problem* remains challenging over subsequent passes. |
BERTops: Studying BERT Representations under a Topological Lens | Proposing scoring functions to effectively understand, analyze and learn
various properties of high dimensional hidden representations of large-scale
transformer models like BERT can be a challenging task. In this work, we
explore a new direction by studying the topological features of BERT hidden
representations using persistent homology (PH). We propose a novel scoring
function named "persistence scoring function (PSF)" which: (i) accurately
captures the homology of the high-dimensional hidden representations and
correlates well with the test set accuracy of a wide range of datasets and
outperforms existing scoring metrics, (ii) captures interesting post
fine-tuning "per-class" level properties from both qualitative and quantitative
viewpoints, (iii) is more stable to perturbations as compared to the baseline
functions, which makes it a very robust proxy, and (iv) finally, also serves as
a predictor of the attack success rates for a wide category of black-box and
white-box adversarial attack methods. Our extensive correlation experiments
demonstrate the practical utility of PSF on various NLP tasks relevant to BERT. |
CCLF: A Contrastive-Curiosity-Driven Learning Framework for Sample-Efficient Reinforcement Learning | In reinforcement learning (RL), it is challenging to learn directly from
high-dimensional observations, where data augmentation has recently been shown
to remedy this via encoding invariances from raw pixels. Nevertheless, we
empirically find that not all samples are equally important and hence simply
injecting more augmented inputs may instead cause instability in Q-learning. In
this paper, we approach this problem systematically by developing a
model-agnostic Contrastive-Curiosity-Driven Learning Framework (CCLF), which
can fully exploit sample importance and improve learning efficiency in a
self-supervised manner. Facilitated by the proposed contrastive curiosity, CCLF
is capable of prioritizing the experience replay, selecting the most
informative augmented inputs, and more importantly regularizing the Q-function
as well as the encoder to concentrate more on under-learned data. Moreover, it
encourages the agent to explore with a curiosity-based reward. As a result, the
agent can focus on more informative samples and learn representation
invariances more efficiently, with significantly reduced augmented inputs. We
apply CCLF to several base RL algorithms and evaluate on the DeepMind Control
Suite, Atari, and MiniGrid benchmarks, where our approach demonstrates superior
sample efficiency and learning performances compared with other
state-of-the-art methods. |
Understanding CNNs from excitations | For instance-level explanation, in order to reveal the relations between
high-level semantics and detailed spatial information, this paper proposes a
novel cognitive approach to neural networks, which named PANE. Under the
guidance of PANE, a novel saliency map representation method, named IOM, is
proposed for CNN-like models. We make the comparison with eight
state-of-the-art saliency map representation methods. The experimental results
show that IOM far outperforms baselines. The work of this paper may bring a new
perspective to understand deep neural networks. |
Revisiting Gaussian Neurons for Online Clustering with Unknown Number of Clusters | Despite the recent success of artificial neural networks, more biologically
plausible learning methods may be needed to resolve the weaknesses of
backpropagation trained models such as catastrophic forgetting and adversarial
attacks. A novel local learning rule is presented that performs online
clustering with a maximum limit of the number of cluster to be found rather
than a fixed cluster count. Instead of using orthogonal weight or output
activation constraints, activation sparsity is achieved by mutual repulsion of
lateral Gaussian neurons ensuring that multiple neuron centers cannot occupy
the same location in the input domain. An update method is also presented for
adjusting the widths of the Gaussian neurons in cases where the data samples
can be represented by means and variances. The algorithms were applied on the
MNIST and CIFAR-10 datasets to create filters capturing the input patterns of
pixel patches of various sizes. The experimental results demonstrate stability
in the learned parameters across a large number of training samples. |
FastGCL: Fast Self-Supervised Learning on Graphs via Contrastive Neighborhood Aggregation | Graph contrastive learning (GCL), as a popular approach to graph
self-supervised learning, has recently achieved a non-negligible effect. To
achieve superior performance, the majority of existing GCL methods elaborate on
graph data augmentation to construct appropriate contrastive pairs. However,
existing methods place more emphasis on the complex graph data augmentation
which requires extra time overhead, and pay less attention to developing
contrastive schemes specific to encoder characteristics. We argue that a better
contrastive scheme should be tailored to the characteristics of graph neural
networks (e.g., neighborhood aggregation) and propose a simple yet effective
method named FastGCL. Specifically, by constructing weighted-aggregated and
non-aggregated neighborhood information as positive and negative samples
respectively, FastGCL identifies the potential semantic information of data
without disturbing the graph topology and node attributes, resulting in faster
training and convergence speeds. Extensive experiments have been conducted on
node classification and graph classification tasks, showing that FastGCL has
competitive classification performance and significant training speedup
compared to existing state-of-the-art methods. |
Positive-Unlabeled Learning with Adversarial Data Augmentation for Knowledge Graph Completion | Most real-world knowledge graphs (KG) are far from complete and
comprehensive. This problem has motivated efforts in predicting the most
plausible missing facts to complete a given KG, i.e., knowledge graph
completion (KGC). However, existing KGC methods suffer from two main issues, 1)
the false negative issue, i.e., the sampled negative training instances may
include potential true facts; and 2) the data sparsity issue, i.e., true facts
account for only a tiny part of all possible facts. To this end, we propose
positive-unlabeled learning with adversarial data augmentation (PUDA) for KGC.
In particular, PUDA tailors positive-unlabeled risk estimator for the KGC task
to deal with the false negative issue. Furthermore, to address the data
sparsity issue, PUDA achieves a data augmentation strategy by unifying
adversarial training and positive-unlabeled learning under the
positive-unlabeled minimax game. Extensive experimental results on real-world
benchmark datasets demonstrate the effectiveness and compatibility of our
proposed method. |
Fast Continuous and Integer L-shaped Heuristics Through Supervised Learning | We propose a methodology at the nexus of operations research and machine
learning (ML) leveraging generic approximators available from ML to accelerate
the solution of mixed-integer linear two-stage stochastic programs. We aim at
solving problems where the second stage is highly demanding. Our core idea is
to gain large reductions in online solution time while incurring small
reductions in first-stage solution accuracy by substituting the exact
second-stage solutions with fast, yet accurate supervised ML predictions. This
upfront investment in ML would be justified when similar problems are solved
repeatedly over time, for example, in transport planning related to fleet
management, routing and container yard management.
Our numerical results focus on the problem class seminally addressed with the
integer and continuous L-shaped cuts. Our extensive empirical analysis is
grounded in standardized families of problems derived from stochastic server
location (SSLP) and stochastic multi knapsack (SMKP) problems available in the
literature. The proposed method can solve the hardest instances of SSLP in less
than 9% of the time it takes the state-of-the-art exact method, and in the case
of SMKP the same figure is 20%. Average optimality gaps are in most cases less
than 0.1%. |
Modeling and mitigation of occupational safety risks in dynamic industrial environments | Identifying and mitigating safety risks is paramount in a number of
industries. In addition to guidelines and best practices, many industries
already have safety management systems (SMSs) designed to monitor and reinforce
good safety behaviors. The analytic capabilities to analyze the data acquired
through such systems, however, are still lacking in terms of their ability to
robustly quantify risks posed by various occupational hazards. Moreover, best
practices and modern SMSs are unable to account for dynamically evolving
environments/behavioral characteristics commonly found in many industrial
settings. This article proposes a method to address these issues by enabling
continuous and quantitative assessment of safety risks in a data-driven manner.
The backbone of our method is an intuitive hierarchical probabilistic model
that explains sparse and noisy safety data collected by a typical SMS. A fully
Bayesian approach is developed to calibrate this model from safety data in an
online fashion. Thereafter, the calibrated model holds necessary information
that serves to characterize risk posed by different safety hazards.
Additionally, the proposed model can be leveraged for automated decision
making, for instance solving resource allocation problems -- targeted towards
risk mitigation -- that are often encountered in resource-constrained
industrial environments. The methodology is rigorously validated on a simulated
test-bed and its scalability is demonstrated on real data from large
maintenance projects at a petrochemical plant. |
WeatherBench Probability: A benchmark dataset for probabilistic medium-range weather forecasting along with deep learning baseline models | WeatherBench is a benchmark dataset for medium-range weather forecasting of
geopotential, temperature and precipitation, consisting of preprocessed data,
predefined evaluation metrics and a number of baseline models. WeatherBench
Probability extends this to probabilistic forecasting by adding a set of
established probabilistic verification metrics (continuous ranked probability
score, spread-skill ratio and rank histograms) and a state-of-the-art
operational baseline using the ECWMF IFS ensemble forecast. In addition, we
test three different probabilistic machine learning methods -- Monte Carlo
dropout, parametric prediction and categorical prediction, in which the
probability distribution is discretized. We find that plain Monte Carlo dropout
severely underestimates uncertainty. The parametric and categorical models both
produce fairly reliable forecasts of similar quality. The parametric models
have fewer degrees of freedom while the categorical model is more flexible when
it comes to predicting non-Gaussian distributions. None of the models are able
to match the skill of the operational IFS model. We hope that this benchmark
will enable other researchers to evaluate their probabilistic approaches. |
Lightweight Image Enhancement Network for Mobile Devices Using Self-Feature Extraction and Dense Modulation | Convolutional neural network (CNN) based image enhancement methods such as
super-resolution and detail enhancement have achieved remarkable performances.
However, amounts of operations including convolution and parameters within the
networks cost high computing power and need huge memory resource, which limits
the applications with on-device requirements. Lightweight image enhancement
network should restore details, texture, and structural information from
low-resolution input images while keeping their fidelity. To address these
issues, a lightweight image enhancement network is proposed. The proposed
network include self-feature extraction module which produces modulation
parameters from low-quality image itself, and provides them to modulate the
features in the network. Also, dense modulation block is proposed for unit
block of the proposed network, which uses dense connections of concatenated
features applied in modulation layers. Experimental results demonstrate better
performance over existing approaches in terms of both quantitative and
qualitative evaluations. |
Model-based Deep Learning Receiver Design for Rate-Splitting Multiple Access | Effective and adaptive interference management is required in next generation
wireless communication systems. To address this challenge, Rate-Splitting
Multiple Access (RSMA), relying on multi-antenna rate-splitting (RS) at the
transmitter and successive interference cancellation (SIC) at the receivers,
has been intensively studied in recent years, albeit mostly under the
assumption of perfect Channel State Information at the Receiver (CSIR) and
ideal capacity-achieving modulation and coding schemes. To assess its practical
performance, benefits, and limits under more realistic conditions, this work
proposes a novel design for a practical RSMA receiver based on model-based deep
learning (MBDL) methods, which aims to unite the simple structure of the
conventional SIC receiver and the robustness and model agnosticism of deep
learning techniques. The MBDL receiver is evaluated in terms of uncoded Symbol
Error Rate (SER), throughput performance through Link-Level Simulations (LLS),
and average training overhead. Also, a comparison with the SIC receiver, with
perfect and imperfect CSIR, is given. Results reveal that the MBDL outperforms
by a significant margin the SIC receiver with imperfect CSIR, due to its
ability to generate on demand non-linear symbol detection boundaries in a pure
data-driven manner. |
Gradient Descent, Stochastic Optimization, and Other Tales | The goal of this paper is to debunk and dispel the magic behind black-box
optimizers and stochastic optimizers. It aims to build a solid foundation on
how and why the techniques work. This manuscript crystallizes this knowledge by
deriving from simple intuitions, the mathematics behind the strategies. This
tutorial doesn't shy away from addressing both the formal and informal aspects
of gradient descent and stochastic optimization methods. By doing so, it hopes
to provide readers with a deeper understanding of these techniques as well as
the when, the how and the why of applying these algorithms.
Gradient descent is one of the most popular algorithms to perform
optimization and by far the most common way to optimize machine learning tasks.
Its stochastic version receives attention in recent years, and this is
particularly true for optimizing deep neural networks. In deep neural networks,
the gradient followed by a single sample or a batch of samples is employed to
save computational resources and escape from saddle points. In 1951, Robbins
and Monro published \textit{A stochastic approximation method}, one of the
first modern treatments on stochastic optimization that estimates local
gradients with a new batch of samples. And now, stochastic optimization has
become a core technology in machine learning, largely due to the development of
the back propagation algorithm in fitting a neural network. The sole aim of
this article is to give a self-contained introduction to concepts and
mathematical tools in gradient descent and stochastic optimization. |
Exploration in Deep Reinforcement Learning: A Survey | This paper reviews exploration techniques in deep reinforcement learning.
Exploration techniques are of primary importance when solving sparse reward
problems. In sparse reward problems, the reward is rare, which means that the
agent will not find the reward often by acting randomly. In such a scenario, it
is challenging for reinforcement learning to learn rewards and actions
association. Thus more sophisticated exploration methods need to be devised.
This review provides a comprehensive overview of existing exploration
approaches, which are categorized based on the key contributions as follows
reward novel states, reward diverse behaviours, goal-based methods,
probabilistic methods, imitation-based methods, safe exploration and
random-based methods. Then, the unsolved challenges are discussed to provide
valuable future research directions. Finally, the approaches of different
categories are compared in terms of complexity, computational effort and
overall performance. |
Deep-Attack over the Deep Reinforcement Learning | Recent adversarial attack developments have made reinforcement learning more
vulnerable, and different approaches exist to deploy attacks against it, where
the key is how to choose the right timing of the attack. Some work tries to
design an attack evaluation function to select critical points that will be
attacked if the value is greater than a certain threshold. This approach makes
it difficult to find the right place to deploy an attack without considering
the long-term impact. In addition, there is a lack of appropriate indicators of
assessment during attacks. To make the attacks more intelligent as well as to
remedy the existing problems, we propose the reinforcement learning-based
attacking framework by considering the effectiveness and stealthy
spontaneously, while we also propose a new metric to evaluate the performance
of the attack model in these two aspects. Experimental results show the
effectiveness of our proposed model and the goodness of our proposed evaluation
metric. Furthermore, we validate the transferability of the model, and also its
robustness under the adversarial training. |
Type-aware Embeddings for Multi-Hop Reasoning over Knowledge Graphs | Multi-hop reasoning over real-life knowledge graphs (KGs) is a highly
challenging problem as traditional subgraph matching methods are not capable to
deal with noise and missing information. To address this problem, it has been
recently introduced a promising approach based on jointly embedding logical
queries and KGs into a low-dimensional space to identify answer entities.
However, existing proposals ignore critical semantic knowledge inherently
available in KGs, such as type information. To leverage type information, we
propose a novel TypE-aware Message Passing (TEMP) model, which enhances the
entity and relation representations in queries, and simultaneously improves
generalization, deductive and inductive reasoning. Remarkably, TEMP is a
plug-and-play model that can be easily incorporated into existing
embedding-based models to improve their performance. Extensive experiments on
three real-world datasets demonstrate TEMP's effectiveness. |
Efficient Accelerator for Dilated and Transposed Convolution with Decomposition | Hardware acceleration for dilated and transposed convolution enables real
time execution of related tasks like segmentation, but current designs are
specific for these convolutional types or suffer from complex control for
reconfigurable designs. This paper presents a design that decomposes input or
weight for dilated and transposed convolutions respectively to skip redundant
computations and thus executes efficiently on existing dense CNN hardware as
well. The proposed architecture can cut down 87.8\% of the cycle counts to
achieve 8.2X speedup over a naive execution for the ENet case. |
Real Time On Sensor Gait Phase Detection with 0.5KB Deep Learning Model | Gait phase detection with convolution neural network provides accurate
classification but demands high computational cost, which inhibits real time
low power on-sensor processing. This paper presents a segmentation based gait
phase detection with a width and depth downscaled U-Net like model that only
needs 0.5KB model size and 67K operations per second with 95.9% accuracy to be
easily fitted into resource limited on sensor microcontroller. |
Pre-RTL DNN Hardware Evaluator With Fused Layer Support | With the popularity of the deep neural network (DNN), hardware accelerators
are demanded for real time execution. However, lengthy design process and fast
evolving DNN models make hardware evaluation hard to meet the time to market
need. This paper proposes a pre-RTL DNN hardware evaluator that supports
conventional layer-by-layer processing as well as the fused layer processing
for low external bandwidth requirement. The evaluator supports two
state-of-the-art accelerator architectures and finds the best hardware and
layer fusion group The experimental results show the layer fusion scheme can
achieve 55.6% memory bandwidth reduction, 36.7% latency improvement and 49.2%
energy reduction compared with layer-by-layer operation. |
A Real Time 1280x720 Object Detection Chip With 585MB/s Memory Traffic | Memory bandwidth has become the real-time bottleneck of current deep learning
accelerators (DLA), particularly for high definition (HD) object detection.
Under resource constraints, this paper proposes a low memory traffic DLA chip
with joint hardware and software optimization. To maximize hardware utilization
under memory bandwidth, we morph and fuse the object detection model into a
group fusion-ready model to reduce intermediate data access. This reduces the
YOLOv2's feature memory traffic from 2.9 GB/s to 0.15 GB/s. To support group
fusion, our previous DLA based hardware employes a unified buffer with
write-masking for simple layer-by-layer processing in a fusion group. When
compared to our previous DLA with the same PE numbers, the chip implemented in
a TSMC 40nm process supports 1280x720@30FPS object detection and consumes 7.9X
less external DRAM access energy, from 2607 mJ to 327.6 mJ. |
PSCNN: A 885.86 TOPS/W Programmable SRAM-based Computing-In-Memory Processor for Keyword Spotting | Computing-in-memory (CIM) has attracted significant attentions in recent
years due to its massive parallelism and low power consumption. However,
current CIM designs suffer from large area overhead of small CIM macros and bad
programmablity for model execution. This paper proposes a programmable CIM
processor with a single large sized CIM macro instead of multiple smaller ones
for power efficient computation and a flexible instruction set to support
various binary 1-D convolution Neural Network (CNN) models in an easy way.
Furthermore, the proposed architecture adopts the pooling write-back method to
support fused or independent convolution/pooling operations to reduce 35.9\% of
latency, and the flexible ping-pong feature SRAM to fit different feature map
sizes during layer-by-layer execution.The design fabricated in TSMC 28nm
technology achieves 150.8 GOPS throughput and 885.86 TOPS/W power efficiency at
10 MHz when executing our binary keyword spotting model, which has higher power
efficiency and flexibility than previous designs. |
RangeSeg: Range-Aware Real Time Segmentation of 3D LiDAR Point Clouds | Semantic outdoor scene understanding based on 3D LiDAR point clouds is a
challenging task for autonomous driving due to the sparse and irregular data
structure. This paper takes advantages of the uneven range distribution of
different LiDAR laser beams to propose a range aware instance segmentation
network, RangeSeg. RangeSeg uses a shared encoder backbone with two range
dependent decoders. A heavy decoder only computes top of a range image where
the far and small objects locate to improve small object detection accuracy,
and a light decoder computes whole range image for low computational cost. The
results are further clustered by the DBSCAN method with a resolution weighted
distance function to get instance-level segmentation results. Experiments on
the KITTI dataset show that RangeSeg outperforms the state-of-the-art semantic
segmentation methods with enormous speedup and improves the instance-level
segmentation performance on small and far objects. The whole RangeSeg pipeline
meets the real time requirement on NVIDIA\textsuperscript{\textregistered}
JETSON AGX Xavier with 19 frames per second in average. |
Zebra: Memory Bandwidth Reduction for CNN Accelerators With Zero Block Regularization of Activation Maps | The large amount of memory bandwidth between local buffer and external DRAM
has become the speedup bottleneck of CNN hardware accelerators, especially for
activation maps. To reduce memory bandwidth, we propose to learn pruning
unimportant blocks dynamically with zero block regularization of activation
maps (Zebra). This strategy has low computational overhead and could easily
integrate with other pruning methods for better performance. The experimental
results show that the proposed method can reduce 70\% of memory bandwidth for
Resnet-18 on Tiny-Imagenet within 1\% accuracy drops and 2\% accuracy gain with
the combination of Network Slimming. |
Sparse Compressed Spiking Neural Network Accelerator for Object Detection | Spiking neural networks (SNNs), which are inspired by the human brain, have
recently gained popularity due to their relatively simple and low-power
hardware for transmitting binary spikes and highly sparse activation maps.
However, because SNNs contain extra time dimension information, the SNN
accelerator will require more buffers and take longer to infer, especially for
the more difficult high-resolution object detection task. As a result, this
paper proposes a sparse compressed spiking neural network accelerator that
takes advantage of the high sparsity of activation maps and weights by
utilizing the proposed gated one-to-all product for low power and highly
parallel model execution. The experimental result of the neural network shows
71.5$\%$ mAP with mixed (1,3) time steps on the IVS 3cls dataset. The
accelerator with the TSMC 28nm CMOS process can achieve 1024$\times$576@29
frames per second processing when running at 500MHz with 35.88TOPS/W energy
efficiency and 1.05mJ energy consumption per frame. |
BSRA: Block-based Super Resolution Accelerator with Hardware Efficient Pixel Attention | Increasingly, convolution neural network (CNN) based super resolution models
have been proposed for better reconstruction results, but their large model
size and complicated structure inhibit their real-time hardware implementation.
Current hardware designs are limited to a plain network and suffer from lower
quality and high memory bandwidth requirements. This paper proposes a super
resolution hardware accelerator with hardware efficient pixel attention that
just needs 25.9K parameters and simple structure but achieves 0.38dB better
reconstruction images than the widely used FSRCNN. The accelerator adopts full
model block wise convolution for full model layer fusion to reduce external
memory access to model input and output only. In addition, CNN and pixel
attention are well supported by PE arrays with distributed weights. The final
implementation can support full HD image reconstruction at 30 frames per second
with TSMC 40nm CMOS process. |
Large Neighborhood Search based on Neural Construction Heuristics | We propose a Large Neighborhood Search (LNS) approach utilizing a learned
construction heuristic based on neural networks as repair operator to solve the
vehicle routing problem with time windows (VRPTW). Our method uses graph neural
networks to encode the problem and auto-regressively decodes a solution and is
trained with reinforcement learning on the construction task without requiring
any labels for supervision. The neural repair operator is combined with a local
search routine, heuristic destruction operators and a selection procedure
applied to a small population to arrive at a sophisticated solution approach.
The key idea is to use the learned model to re-construct the partially
destructed solution and to introduce randomness via the destruction heuristics
(or the stochastic policy itself) to effectively explore a large neighborhood. |
Data-driven emotional body language generation for social robotics | In social robotics, endowing humanoid robots with the ability to generate
bodily expressions of affect can improve human-robot interaction and
collaboration, since humans attribute, and perhaps subconsciously anticipate,
such traces to perceive an agent as engaging, trustworthy, and socially
present. Robotic emotional body language needs to be believable, nuanced and
relevant to the context. We implemented a deep learning data-driven framework
that learns from a few hand-designed robotic bodily expressions and can
generate numerous new ones of similar believability and lifelikeness. The
framework uses the Conditional Variational Autoencoder model and a sampling
approach based on the geometric properties of the model's latent space to
condition the generative process on targeted levels of valence and arousal. The
evaluation study found that the anthropomorphism and animacy of the generated
expressions are not perceived differently from the hand-designed ones, and the
emotional conditioning was adequately differentiable between most levels except
the pairs of neutral-positive valence and low-medium arousal. Furthermore, an
exploratory analysis of the results reveals a possible impact of the
conditioning on the perceived dominance of the robot, as well as on the
participants' attention. |
VICE: Variational Interpretable Concept Embeddings | A central goal in the cognitive sciences is the development of numerical
models for mental representations of object concepts. This paper introduces
Variational Interpretable Concept Embeddings (VICE), an approximate Bayesian
method for embedding object concepts in a vector space using data collected
from humans in an odd-one-out triplet task. VICE uses variational inference to
obtain sparse, non-negative representations of object concepts with uncertainty
estimates for the embedding values. These estimates are used to automatically
select the dimensions that best explain the data. We derive a PAC learning
bound for VICE that can be used to estimate generalization performance or
determine sufficient sample size in experimental design. VICE rivals or
outperforms its predecessor, SPoSE, at predicting human behavior in the
odd-one-out triplet task. Furthermore, VICE's object representations are more
reproducible and consistent across random initializations. |
Smoothed Online Convex Optimization Based on Discounted-Normal-Predictor | In this paper, we investigate an online prediction strategy named as
Discounted-Normal-Predictor (Kapralov and Panigrahy, 2010) for smoothed online
convex optimization (SOCO), in which the learner needs to minimize not only the
hitting cost but also the switching cost. In the setting of learning with
expert advice, Daniely and Mansour (2019) demonstrate that
Discounted-Normal-Predictor can be utilized to yield nearly optimal regret
bounds over any interval, even in the presence of switching costs. Inspired by
their results, we develop a simple algorithm for SOCO: Combining online
gradient descent (OGD) with different step sizes sequentially by
Discounted-Normal-Predictor. Despite its simplicity, we prove that it is able
to minimize the adaptive regret with switching cost, i.e., attaining nearly
optimal regret with switching cost on every interval. By exploiting the
theoretical guarantee of OGD for dynamic regret, we further show that the
proposed algorithm can minimize the dynamic regret with switching cost in every
interval. |
FedDKD: Federated Learning with Decentralized Knowledge Distillation | The performance of federated learning in neural networks is generally
influenced by the heterogeneity of the data distribution. For a well-performing
global model, taking a weighted average of the local models, as done by most
existing federated learning algorithms, may not guarantee consistency with
local models in the space of neural network maps. In this paper, we propose a
novel framework of federated learning equipped with the process of
decentralized knowledge distillation (FedDKD) (i.e., without data on the
server). The FedDKD introduces a module of decentralized knowledge distillation
(DKD) to distill the knowledge of the local models to train the global model by
approaching the neural network map average based on the metric of divergence
defined in the loss function, other than only averaging parameters as done in
literature. Numeric experiments on various heterogeneous datasets reveal that
FedDKD outperforms the state-of-the-art methods with more efficient
communication and training in a few DKD steps, especially on some extremely
heterogeneous datasets. |
DeepGraviLens: a Multi-Modal Architecture for Classifying Gravitational Lensing Data | Gravitational lensing is the relativistic effect generated by massive bodies,
which bend the space-time surrounding them. It is a deeply investigated topic
in astrophysics and allows validating theoretical relativistic results and
studying faint astrophysical objects that would not be visible otherwise. In
recent years Machine Learning methods have been applied to support the analysis
of the gravitational lensing phenomena by detecting lensing effects in data
sets consisting of images associated with brightness variation time series.
However, the state-of-art approaches either consider only images and neglect
time-series data or achieve relatively low accuracy on the most difficult data
sets. This paper introduces DeepGraviLens, a novel multi-modal network that
classifies spatio-temporal data belonging to one non-lensed system type and
three lensed system types. It surpasses the current state of the art accuracy
results by $\approx$ 19% to $\approx$ 43%, depending on the considered data
set. Such an improvement will enable the acceleration of the analysis of lensed
objects in upcoming astrophysical surveys, which will exploit the petabytes of
data collected, e.g., from the Vera C. Rubin Observatory. |
Unsupervised Denoising of Optical Coherence Tomography Images with Dual_Merged CycleWGAN | Nosie is an important cause of low quality Optical coherence tomography (OCT)
image. The neural network model based on Convolutional neural networks(CNNs)
has demonstrated its excellent performance in image denoising. However, OCT
image denoising still faces great challenges because many previous neural
network algorithms required a large number of labeled data, which might cost
much time or is expensive. Besides, these CNN-based algorithms need numerous
parameters and good tuning techniques, which is hardware resources consuming.
To solved above problems, We proposed a new Cycle-Consistent Generative
Adversarial Nets called Dual-Merged Cycle-WGAN for retinal OCT image
denoiseing, which has remarkable performance with less unlabeled traning data.
Our model consists of two Cycle-GAN networks with imporved generator,
descriminator and wasserstein loss to achieve good training stability and
better performance. Using image merge technique between two Cycle-GAN networks,
our model could obtain more detailed information and hence better training
effect. The effectiveness and generality of our proposed network has been
proved via ablation experiments and comparative experiments. Compared with
other state-of-the-art methods, our unsupervised method obtains best subjective
visual effect and higher evaluation objective indicators. |
A Multi-stage deep architecture for summary generation of soccer videos | Video content is present in an ever-increasing number of fields, both
scientific and commercial. Sports, particularly soccer, is one of the
industries that has invested the most in the field of video analytics, due to
the massive popularity of the game and the emergence of new markets. Previous
state-of-the-art methods on soccer matches video summarization rely on
handcrafted heuristics to generate summaries which are poorly generalizable,
but these works have yet proven that multiple modalities help detect the best
actions of the game. On the other hand, machine learning models with higher
generalization potential have entered the field of summarization of
general-purpose videos, offering several deep learning approaches. However,
most of them exploit content specificities that are not appropriate for sport
whole-match videos. Although video content has been for many years the main
source for automatizing knowledge extraction in soccer, the data that records
all the events happening on the field has become lately very important in
sports analytics, since this event data provides richer context information and
requires less processing. We propose a method to generate the summary of a
soccer match exploiting both the audio and the event metadata. The results show
that our method can detect the actions of the match, identify which of these
actions should belong to the summary and then propose multiple candidate
summaries which are similar enough but with relevant variability to provide
different options to the final editor. Furthermore, we show the generalization
capability of our work since it can transfer knowledge between datasets from
different broadcasting companies, different competitions, acquired in different
conditions, and corresponding to summaries of different lengths |
From Noisy Prediction to True Label: Noisy Prediction Calibration via Generative Model | Noisy labels are inevitable yet problematic in machine learning society. It
ruins the generalization of a classifier by making the classifier over-fitted
to noisy labels. Existing methods on noisy label have focused on modifying the
classifier during the training procedure. It has two potential problems. First,
these methods are not applicable to a pre-trained classifier without further
access to training. Second, it is not easy to train a classifier and regularize
all negative effects from noisy labels, simultaneously. We suggest a new branch
of method, Noisy Prediction Calibration (NPC) in learning with noisy labels.
Through the introduction and estimation of a new type of transition matrix via
generative model, NPC corrects the noisy prediction from the pre-trained
classifier to the true label as a post-processing scheme. We prove that NPC
theoretically aligns with the transition matrix based methods. Yet, NPC
empirically provides more accurate pathway to estimate true label, even without
involvement in classifier learning. Also, NPC is applicable to any classifier
trained with noisy label methods, if training instances and its predictions are
available. Our method, NPC, boosts the classification performances of all
baseline models on both synthetic and real-world datasets. The implemented code
is available at https://github.com/BaeHeeSun/NPC. |
Simple Techniques Work Surprisingly Well for Neural Network Test Prioritization and Active Learning (Replicability Study) | Test Input Prioritizers (TIP) for Deep Neural Networks (DNN) are an important
technique to handle the typically very large test datasets efficiently, saving
computation and labeling costs. This is particularly true for large-scale,
deployed systems, where inputs observed in production are recorded to serve as
potential test or training data for the next versions of the system. Feng et.
al. propose DeepGini, a very fast and simple TIP, and show that it outperforms
more elaborate techniques such as neuron- and surprise coverage. In a
large-scale study (4 case studies, 8 test datasets, 32'200 trained models) we
verify their findings. However, we also find that other comparable or even
simpler baselines from the field of uncertainty quantification, such as the
predicted softmax likelihood or the entropy of the predicted softmax
likelihoods perform equally well as DeepGini. |
Skeptical binary inferences in multi-label problems with sets of probabilities | In this paper, we consider the problem of making distributionally robust,
skeptical inferences for the multi-label problem, or more generally for Boolean
vectors. By distributionally robust, we mean that we consider a set of possible
probability distributions, and by skeptical we understand that we consider as
valid only those inferences that are true for every distribution within this
set. Such inferences will provide partial predictions whenever the considered
set is sufficiently big. We study in particular the Hamming loss case, a common
loss function in multi-label problems, showing how skeptical inferences can be
made in this setting. Our experimental results are organised in three sections;
(1) the first one indicates the gain computational obtained from our
theoretical results by using synthetical data sets, (2) the second one
indicates that our approaches produce relevant cautiousness on those
hard-to-predict instances where its precise counterpart fails, and (3) the last
one demonstrates experimentally how our approach copes with imperfect
information (generated by a downsampling procedure) better than the partial
abstention [31] and the rejection rules. |
The Multivariate Community Hawkes Model for Dependent Relational Events in Continuous-time Networks | The stochastic block model (SBM) is one of the most widely used generative
models for network data. Many continuous-time dynamic network models are built
upon the same assumption as the SBM: edges or events between all pairs of nodes
are conditionally independent given the block or community memberships, which
prevents them from reproducing higher-order motifs such as triangles that are
commonly observed in real networks. We propose the multivariate community
Hawkes (MULCH) model, an extremely flexible community-based model for
continuous-time networks that introduces dependence between node pairs using
structured multivariate Hawkes processes. We fit the model using a spectral
clustering and likelihood-based local refinement procedure. We find that our
proposed MULCH model is far more accurate than existing models both for
predictive and generative tasks. |
PSI Draft Specification | This document presents the draft specification for delivering machine
learning services over HTTP, developed as part of the Protocols and Structures
for Inference project, which concluded in 2013. It presents the motivation for
providing machine learning as a service, followed by a description of the
essential and optional components of such a service. |
Community detection in multiplex networks based on orthogonal nonnegative matrix tri-factorization | Networks provide a powerful tool to model complex systems where the different
entities in the system are presented by nodes and their interactions by edges.
Recently, there has been a growing interest in multiplex networks as they can
represent the interactions between a pair of nodes through multiple types of
links, each reflecting a distinct type of interaction. One of the important
tools in understanding network topology is community detection. Although there
are numerous works on community detection in single layer networks, existing
work on multiplex community detection mostly focuses on learning a common
community structure across layers without taking the heterogeneity of the
different layers into account. In this paper, we introduce a new multiplex
community detection approach that can identify communities that are common
across layers as well as those that are unique to each layer. The proposed
algorithm employs Orthogonal Nonnegative Matrix Tri-Factorization to model each
layer's adjacency matrix as the sum of two low-rank matrix factorizations,
corresponding to the common and private communities, respectively. The proposed
algorithm is evaluated on both synthetic and real multiplex networks and
compared to state-of-the-art techniques. |
LoopStack: a Lightweight Tensor Algebra Compiler Stack | We present LoopStack, a domain specific compiler stack for tensor operations,
composed of a frontend, LoopTool, and an efficient optimizing code generator,
LoopNest. This stack enables us to compile entire neural networks and generate
code targeting the AVX2, AVX512, NEON, and NEONfp16 instruction sets while
incorporating optimizations often missing from other machine learning compiler
backends. We evaluate our stack on a collection of full neural networks and
commonly used network blocks as well as individual operators, and show that
LoopStack generates machine code that matches and frequently exceeds the
performance of in state-of-the-art machine learning frameworks in both cases.
We also show that for a large collection of schedules LoopNest's compilation is
orders of magnitude faster than LLVM, while resulting in equal or improved run
time performance. Additionally, LoopStack has a very small memory footprint - a
binary size of 245KB, and under 30K lines of effective code makes it ideal for
use on mobile and embedded devices. |
Physics-aware Reduced-order Modeling of Transonic Flow via $β$-Variational Autoencoder | Autoencoder-based reduced-order modeling (ROM) has recently attracted
significant attention, owing to its ability to capture underlying nonlinear
features. However, two critical drawbacks severely undermine its scalability to
various physical applications: entangled and therefore uninterpretable latent
variables (LVs) and the blindfold determination of latent space dimension. In
this regard, this study proposes the physics-aware ROM using only interpretable
and information-intensive LVs extracted by $\beta$-variational autoencoder,
which are referred to as physics-aware LVs throughout this paper. To extract
these LVs, their independence and information intensity are quantitatively
scrutinized in a two-dimensional transonic flow benchmark problem. Then, the
physical meanings of the physics-aware LVs are thoroughly investigated and we
confirmed that with appropriate hyperparameter $\beta$, they actually
correspond to the generating factors of the training dataset, Mach number and
angle of attack. To the best of the authors' knowledge, our work is the first
to practically confirm that $\beta$-variational autoencoder can automatically
extract the physical generating factors in the field of applied physics.
Finally, physics-aware ROM, which utilizes only physics-aware LVs, is compared
with conventional ROMs, and its validity and efficiency are successfully
verified. |
Forecasting Market Changes using Variational Inference | Though various approaches have been considered, forecasting near-term market
changes of equities and similar market data remains quite difficult. In this
paper we introduce an approach to forecast near-term market changes for equity
indices as well as portfolios using variational inference (VI). VI is a machine
learning approach which uses optimization techniques to estimate complex
probability densities. In the proposed approach, clusters of explanatory
variables are identified and market changes are forecast based on
cluster-specific linear regression. Apart from the expected value of changes,
the proposed approach can also be used to obtain the distribution of possible
outcomes, which can be used to estimate confidence levels of forecasts and risk
measures such as VaR (Value at Risk) for the portfolio. Another advantage of
the proposed approach is the clear model interpretation, as clusters of
explanatory variables (or market regimes) are identified for which the future
changes follow similar relationships. Knowledge about such clusters can provide
useful insights about portfolio performance and identify the relative
importance of variables in different market regimes. Illustrative examples of
equity and bond indices are considered to demonstrate forecasts of the proposed
approach during Covid-related volatility in early 2020 and subsequent benign
market conditions. For the portfolios considered, it is shown that the proposed
approach provides useful forecasts in both normal and volatile markets even
with only a few explanatory variables. Additionally the predicted estimate and
distribution adapt quickly to changing market conditions and thus may also be
useful in obtaining better real-time estimates of risk measures such as VaR
compared to traditional approaches. |
Attention-wise masked graph contrastive learning for predicting molecular property | Accurate and efficient prediction of the molecular properties of drugs is one
of the fundamental problems in drug research and development. Recent
advancements in representation learning have been shown to greatly improve the
performance of molecular property prediction. However, due to limited labeled
data, supervised learning-based molecular representation algorithms can only
search limited chemical space, which results in poor generalizability. In this
work, we proposed a self-supervised representation learning framework for
large-scale unlabeled molecules. We developed a novel molecular graph
augmentation strategy, referred to as attention-wise graph mask, to generate
challenging positive sample for contrastive learning. We adopted the graph
attention network (GAT) as the molecular graph encoder, and leveraged the
learned attention scores as masking guidance to generate molecular augmentation
graphs. By minimization of the contrastive loss between original graph and
masked graph, our model can capture important molecular structure and
higher-order semantic information. Extensive experiments showed that our
attention-wise graph mask contrastive learning exhibit state-of-the-art
performance in a couple of downstream molecular property prediction tasks. |
Using a novel fractional-order gradient method for CNN back-propagation | Computer-aided diagnosis tools have experienced rapid growth and development
in recent years. Among all, deep learning is the most sophisticated and popular
tool. In this paper, researchers propose a novel deep learning model and apply
it to COVID-19 diagnosis. Our model uses the tool of fractional calculus, which
has the potential to improve the performance of gradient methods. To this end,
the researcher proposes a fractional-order gradient method for the
back-propagation of convolutional neural networks based on the Caputo
definition. However, if only the first term of the infinite series of the
Caputo definition is used to approximate the fractional-order derivative, the
length of the memory is truncated. Therefore, the fractional-order gradient
(FGD) method with a fixed memory step and an adjustable number of terms is used
to update the weights of the layers. Experiments were performed on the COVIDx
dataset to demonstrate fast convergence, good accuracy, and the ability to
bypass the local optimal point. We also compared the performance of the
developed fractional-order neural networks and Integer-order neural networks.
The results confirmed the effectiveness of our proposed model in the diagnosis
of COVID-19. |
Data-driven control of spatiotemporal chaos with reduced-order neural ODE-based models and reinforcement learning | Deep reinforcement learning (RL) is a data-driven method capable of
discovering complex control strategies for high-dimensional systems, making it
promising for flow control applications. In particular, the present work is
motivated by the goal of reducing energy dissipation in turbulent flows, and
the example considered is the spatiotemporally chaotic dynamics of the
Kuramoto-Sivashinsky equation (KSE). A major challenge associated with RL is
that substantial training data must be generated by repeatedly interacting with
the target system, making it costly when the system is computationally or
experimentally expensive. We mitigate this challenge in a data-driven manner by
combining dimensionality reduction via an autoencoder with a neural ODE
framework to obtain a low-dimensional dynamical model from just a limited data
set. We substitute this data-driven reduced-order model (ROM) in place of the
true system during RL training to efficiently estimate the optimal policy,
which can then be deployed on the true system. For the KSE actuated with
localized forcing ("jets") at four locations, we demonstrate that we are able
to learn a ROM that accurately captures the actuated dynamics as well as the
underlying natural dynamics just from snapshots of the KSE experiencing random
actuations. Using this ROM and a control objective of minimizing dissipation
and power cost, we extract a control policy from it using deep RL. We show that
the ROM-based control strategy translates well to the true KSE and highlight
that the RL agent discovers and stabilizes an underlying forced equilibrium
solution of the KSE system. We show that this forced equilibrium captured in
the ROM and discovered through RL is related to an existing known equilibrium
solution of the natural KSE. |
Thermodynamically Consistent Machine-Learned Internal State Variable Approach for Data-Driven Modeling of Path-Dependent Materials | Characterization and modeling of path-dependent behaviors of complex
materials by phenomenological models remains challenging due to difficulties in
formulating mathematical expressions and internal state variables (ISVs)
governing path-dependent behaviors. Data-driven machine learning models, such
as deep neural networks and recurrent neural networks (RNNs), have become
viable alternatives. However, pure black-box data-driven models mapping inputs
to outputs without considering the underlying physics suffer from unstable and
inaccurate generalization performance. This study proposes a machine-learned
physics-informed data-driven constitutive modeling approach for path-dependent
materials based on the measurable material states. The proposed data-driven
constitutive model is designed with the consideration of universal
thermodynamics principles, where the ISVs essential to the material
path-dependency are inferred automatically from the hidden state of RNNs. The
RNN describing the evolution of the data-driven machine-learned ISVs follows
the thermodynamics second law. To enhance the robustness and accuracy of RNN
models, stochasticity is introduced to model training. The effects of the
number of RNN history steps, the internal state dimension, the model
complexity, and the strain increment on model performances have been
investigated. The effectiveness of the proposed method is evaluated by modeling
soil material behaviors under cyclic shear loading using experimental
stress-strain data. |
Medical Coding with Biomedical Transformer Ensembles and Zero/Few-shot Learning | Medical coding (MC) is an essential pre-requisite for reliable data retrieval
and reporting. Given a free-text reported term (RT) such as "pain of right
thigh to the knee", the task is to identify the matching lowest-level term
(LLT) - in this case "unilateral leg pain" - from a very large and continuously
growing repository of standardized medical terms. However, automating this task
is challenging due to a large number of LLT codes (as of writing over 80,000),
limited availability of training data for long tail/emerging classes, and the
general high accuracy demands of the medical domain. With this paper, we
introduce the MC task, discuss its challenges, and present a novel approach
called xTARS that combines traditional BERT-based classification with a recent
zero/few-shot learning approach (TARS). We present extensive experiments that
show that our combined approach outperforms strong baselines, especially in the
few-shot regime. The approach is developed and deployed at Bayer, live since
November 2021. As we believe our approach potentially promising beyond MC, and
to ensure reproducibility, we release the code to the research community. |
Experimental quantum pattern recognition in IBMQ and diamond NVs | One of the most promising applications of quantum computing is the processing
of graphical data like images. Here, we investigate the possibility of
realizing a quantum pattern recognition protocol based on swap test, and use
the IBMQ noisy intermediate-scale quantum (NISQ) devices to verify the idea. We
find that with a two-qubit protocol, swap test can efficiently detect the
similarity between two patterns with good fidelity, though for three or more
qubits the noise in the real devices becomes detrimental. To mitigate this
noise effect, we resort to destructive swap test, which shows an improved
performance for three-qubit states. Due to limited cloud access to larger IBMQ
processors, we take a segment-wise approach to apply the destructive swap test
on higher dimensional images. In this case, we define an average overlap
measure which shows faithfulness to distinguish between two very different or
very similar patterns when simulated on real IBMQ processors. As test images,
we use binary images with simple patterns, greyscale MNIST numbers and MNIST
fashion images, as well as binary images of human blood vessel obtained from
magnetic resonance imaging (MRI). We also present an experimental set up for
applying destructive swap test using the nitrogen vacancy centre (NVs) in
diamond. Our experimental data show high fidelity for single qubit states.
Lastly, we propose a protocol inspired from quantum associative memory, which
works in an analogous way to supervised learning for performing quantum pattern
recognition using destructive swap test. |
Federated Semi-Supervised Classification of Multimedia Flows for 3D Networks | Automatic traffic classification is increasingly becoming important in
traffic engineering, as the current trend of encrypting transport information
(e.g., behind HTTP-encrypted tunnels) prevents intermediate nodes from
accessing end-to-end packet headers. However, this information is crucial for
traffic shaping, network slicing, and Quality of Service (QoS) management, for
preventing network intrusion, and for anomaly detection. 3D networks offer
multiple routes that can guarantee different levels of QoS. Therefore, service
classification and separation are essential to guarantee the required QoS level
to each traffic sub-flow through the appropriate network trunk. In this paper,
a federated feature selection and feature reduction learning scheme is proposed
to classify network traffic in a semi-supervised cooperative manner. The
federated gateways of 3D network help to enhance the global knowledge of
network traffic to improve the accuracy of anomaly and intrusion detection and
service identification of a new traffic flow. |
Can Information Behaviour Inform Machine Learning? | The objective of this paper is to explore the opportunities for human
information behaviour research to inform and influence the field of machine
learning and the resulting machine information behaviour. Using the development
of foundation models in machine learning as an example, the paper illustrates
how human information behaviour research can bring to machine learning a more
nuanced view of information and informing, a better understanding of
information need and how that affects the communication among people and
systems, guidance on the nature of context and how to operationalize that in
models and systems, and insights into bias, misinformation, and
marginalization. Despite their clear differences, the fields of information
behaviour and machine learning share many common objectives, paradigms, and key
research questions. The example of foundation models illustrates that human
information behaviour research has much to offer in addressing some of the
challenges emerging in the nascent area of machine information behaviour. |
Generalized Reference Kernel for One-class Classification | In this paper, we formulate a new generalized reference kernel hoping to
improve the original base kernel using a set of reference vectors. Depending on
the selected reference vectors, our formulation shows similarities to
approximate kernels, random mappings, and Non-linear Projection Trick. Focusing
on small-scale one-class classification, our analysis and experimental results
show that the new formulation provides approaches to regularize, adjust the
rank, and incorporate additional information into the kernel itself, leading to
improved one-class classification accuracy. |
Deep vs. Shallow Learning: A Benchmark Study in Low Magnitude Earthquake Detection | While deep learning models have seen recent high uptake in the geosciences,
and are appealing in their ability to learn from minimally processed input
data, as black box models they do not provide an easy means to understand how a
decision is reached, which in safety-critical tasks especially can be
problematical. An alternative route is to use simpler, more transparent white
box models, in which task-specific feature construction replaces the more
opaque feature discovery process performed automatically within deep learning
models. Using data from the Groningen Gas Field in the Netherlands, we build on
an existing logistic regression model by the addition of four further features
discovered using elastic net driven data mining within the catch22 time series
analysis package. We then evaluate the performance of the augmented logistic
regression model relative to a deep (CNN) model, pre-trained on the Groningen
data, on progressively increasing noise-to-signal ratios. We discover that, for
each ratio, our logistic regression model correctly detects every earthquake,
while the deep model fails to detect nearly 20 % of seismic events, thus
justifying at least a degree of caution in the application of deep models,
especially to data with higher noise-to-signal ratios. |
Deep Learning with Logical Constraints | In recent years, there has been an increasing interest in exploiting
logically specified background knowledge in order to obtain neural models (i)
with a better performance, (ii) able to learn from less data, and/or (iii)
guaranteed to be compliant with the background knowledge itself, e.g., for
safety-critical applications. In this survey, we retrace such works and
categorize them based on (i) the logical language that they use to express the
background knowledge and (ii) the goals that they achieve. |
Conditional $β$-VAE for De Novo Molecular Generation | Deep learning has significantly advanced and accelerated de novo molecular
generation. Generative networks, namely Variational Autoencoders (VAEs) can not
only randomly generate new molecules, but also alter molecular structures to
optimize specific chemical properties which are pivotal for drug-discovery.
While VAEs have been proposed and researched in the past for pharmaceutical
applications, they possess deficiencies which limit their ability to both
optimize properties and decode syntactically valid molecules. We present a
recurrent, conditional $\beta$-VAE which disentangles the latent space to
enhance post hoc molecule optimization. We create a mutual information driven
training protocol and data augmentations to both increase molecular validity
and promote longer sequence generation. We demonstrate the efficacy of our
framework on the ZINC-250k dataset, achieving SOTA unconstrained optimization
results on the penalized LogP (pLogP) and QED scores, while also matching
current SOTA results for validity, novelty and uniqueness scores for random
generation. We match the current SOTA on QED for top-3 molecules at 0.948,
while setting a new SOTA for pLogP optimization at 104.29, 90.12, 69.68 and
demonstrating improved results on the constrained optimization task. |
Accurate non-stationary short-term traffic flow prediction method | Precise and timely traffic flow prediction plays a critical role in
developing intelligent transportation systems and has attracted considerable
attention in recent decades. Despite the significant progress in this area
brought by deep learning, challenges remain. Traffic flows usually change
dramatically in a short period, which prevents the current methods from
accurately capturing the future trend and likely causes the over-fitting
problem, leading to unsatisfied accuracy. To this end, this paper proposes a
Long Short-Term Memory (LSTM) based method that can forecast the short-term
traffic flow precisely and avoid local optimum problems during training.
Specifically, instead of using the non-stationary raw traffic data directly, we
first decompose them into sub-components, where each one is less noisy than the
original input. Afterward, Sample Entropy (SE) is employed to merge similar
components to reduce the computation cost. The merged features are fed into the
LSTM, and we then introduce a spatiotemporal module to consider the neighboring
relationships in the recombined signals to avoid strong autocorrelation. During
training, we utilize the Grey Wolf Algorithm (GWO) to optimize the parameters
of LSTM, which overcome the overfitting issue. We conduct the experiments on a
UK public highway traffic flow dataset, and the results show that the proposed
method performs favorably against other state-of-the-art methods with better
adaption performance on extreme outliers, delay effects, and trend-changing
responses. |
An Early Fault Detection Method of Rotating Machines Based on Multiple Feature Fusion with Stacking Architecture | Early fault detection (EFD) of rotating machines is important to decrease the
maintenance cost and improve the mechanical system stability. One of the key
points of EFD is developing a generic model to extract robust and
discriminative features from different equipment for early fault detection.
Most existing EFD methods focus on learning fault representation by one type of
feature. However, a combination of multiple features can capture a more
comprehensive representation of system state. In this paper, we propose an EFD
method based on multiple feature fusion with stacking architecture (M2FSA). The
proposed method can extract generic and discriminiative features to detect
early faults by combining time domain (TD), frequency domain (FD), and
time-frequency domain (TFD) features. In order to unify the dimensions of the
different domain features, Stacked Denoising Autoencoder (SDAE) is utilized to
learn deep features in three domains. The architecture of the proposed M2FSA
consists of two layers. The first layer contains three base models, whose
corresponding inputs are different deep features. The outputs of the first
layer are concatenated to generate the input to the second layer, which
consists of a meta model. The proposed method is tested on three bearing
datasets. The results demonstrate that the proposed method is better than
existing methods both in sensibility and reliability. |
Preserve Pre-trained Knowledge: Transfer Learning With Self-Distillation For Action Recognition | Video-based action recognition is one of the most popular topics in computer
vision. With recent advances of selfsupervised video representation learning
approaches, action recognition usually follows a two-stage training framework,
i.e., self-supervised pre-training on large-scale unlabeled sets and transfer
learning on a downstream labeled set. However, catastrophic forgetting of the
pre-trained knowledge becomes the main issue in the downstream transfer
learning of action recognition, resulting in a sub-optimal solution. In this
paper, to alleviate the above issue, we propose a novel transfer learning
approach that combines self-distillation in fine-tuning to preserve knowledge
from the pre-trained model learned from the large-scale dataset. Specifically,
we fix the encoder from the last epoch as the teacher model to guide the
training of the encoder from the current epoch in the transfer learning. With
such a simple yet effective learning strategy, we outperform state-of-the-art
methods on widely used UCF101 and HMDB51 datasets in action recognition task. |
Domain Adaptation meets Individual Fairness. And they get along | Many instances of algorithmic bias are caused by distributional shifts. For
example, machine learning (ML) models often perform worse on demographic groups
that are underrepresented in the training data. In this paper, we leverage this
connection between algorithmic fairness and distribution shifts to show that
algorithmic fairness interventions can help ML models overcome distribution
shifts, and that domain adaptation methods (for overcoming distribution shifts)
can mitigate algorithmic biases. In particular, we show that (i) enforcing
suitable notions of individual fairness (IF) can improve the
out-of-distribution accuracy of ML models, and that (ii) it is possible to
adapt representation alignment methods for domain adaptation to enforce
(individual) fairness. The former is unexpected because IF interventions were
not developed with distribution shifts in mind. The latter is also unexpected
because representation alignment is not a common approach in the IF literature. |
Is Your Toxicity My Toxicity? Exploring the Impact of Rater Identity on Toxicity Annotation | Machine learning models are commonly used to detect toxicity in online
conversations. These models are trained on datasets annotated by human raters.
We explore how raters' self-described identities impact how they annotate
toxicity in online comments. We first define the concept of specialized rater
pools: rater pools formed based on raters' self-described identities, rather
than at random. We formed three such rater pools for this study--specialized
rater pools of raters from the U.S. who identify as African American, LGBTQ,
and those who identify as neither. Each of these rater pools annotated the same
set of comments, which contains many references to these identity groups. We
found that rater identity is a statistically significant factor in how raters
will annotate toxicity for identity-related annotations. Using preliminary
content analysis, we examined the comments with the most disagreement between
rater pools and found nuanced differences in the toxicity annotations. Next, we
trained models on the annotations from each of the different rater pools, and
compared the scores of these models on comments from several test sets.
Finally, we discuss how using raters that self-identify with the subjects of
comments can create more inclusive machine learning models, and provide more
nuanced ratings than those by random raters. |
On the speed of uniform convergence in Mercer's theorem | The classical Mercer's theorem claims that a continuous positive definite
kernel $K({\mathbf x}, {\mathbf y})$ on a compact set can be represented as
$\sum_{i=1}^\infty \lambda_i\phi_i({\mathbf x})\phi_i({\mathbf y})$ where
$\{(\lambda_i,\phi_i)\}$ are eigenvalue-eigenvector pairs of the corresponding
integral operator. This infinite representation is known to converge uniformly
to the kernel $K$. We estimate the speed of this convergence in terms of the
decay rate of eigenvalues and demonstrate that for $3m$ times differentiable
kernels the first $N$ terms of the series approximate $K$ as
$\mathcal{O}\big((\sum_{i=N+1}^\infty\lambda_i)^{\frac{m}{m+n}}\big)$ or
$\mathcal{O}\big((\sum_{i=N+1}^\infty\lambda^2_i)^{\frac{m}{2m+n}}\big)$. |
Dynamic Programming in Rank Space: Scaling Structured Inference with Low-Rank HMMs and PCFGs | Hidden Markov Models (HMMs) and Probabilistic Context-Free Grammars (PCFGs)
are widely used structured models, both of which can be represented as factor
graph grammars (FGGs), a powerful formalism capable of describing a wide range
of models. Recent research found it beneficial to use large state spaces for
HMMs and PCFGs. However, inference with large state spaces is computationally
demanding, especially for PCFGs. To tackle this challenge, we leverage tensor
rank decomposition (aka.\ CPD) to decrease inference computational complexities
for a subset of FGGs subsuming HMMs and PCFGs. We apply CPD on the factors of
an FGG and then construct a new FGG defined in the rank space. Inference with
the new FGG produces the same result but has a lower time complexity when the
rank size is smaller than the state size. We conduct experiments on HMM
language modeling and unsupervised PCFG parsing, showing better performance
than previous work. Our code is publicly available at
\url{https://github.com/VPeterV/RankSpace-Models}. |
Ridgeless Regression with Random Features | Recent theoretical studies illustrated that kernel ridgeless regression can
guarantee good generalization ability without an explicit regularization. In
this paper, we investigate the statistical properties of ridgeless regression
with random features and stochastic gradient descent. We explore the effect of
factors in the stochastic gradient and random features, respectively.
Specifically, random features error exhibits the double-descent curve.
Motivated by the theoretical findings, we propose a tunable kernel algorithm
that optimizes the spectral density of kernel during training. Our work bridges
the interpolation theory and practical algorithm. |
None Class Ranking Loss for Document-Level Relation Extraction | Document-level relation extraction (RE) aims at extracting relations among
entities expressed across multiple sentences, which can be viewed as a
multi-label classification problem. In a typical document, most entity pairs do
not express any pre-defined relation and are labeled as "none" or "no
relation". For good document-level RE performance, it is crucial to distinguish
such none class instances (entity pairs) from those of pre-defined classes
(relations). However, most existing methods only estimate the probability of
pre-defined relations independently without considering the probability of "no
relation". This ignores the context of entity pairs and the label correlations
between the none class and pre-defined classes, leading to sub-optimal
predictions. To address this problem, we propose a new multi-label loss that
encourages large margins of label confidence scores between each pre-defined
class and the none class, which enables captured label correlations and
context-dependent thresholding for label prediction. To gain further robustness
against positive-negative imbalance and mislabeled data that could appear in
real-world RE datasets, we propose a margin regularization and a margin
shifting technique. Experimental results demonstrate that our method
significantly outperforms existing multi-label losses for document-level RE and
works well in other multi-label tasks such as emotion classification when none
class instances are available for training. |
A Survey of Decentralized Online Learning | Decentralized online learning (DOL) has been increasingly researched in the
last decade, mostly motivated by its wide applications in sensor networks,
commercial buildings, robotics (e.g., decentralized target tracking and
formation control), smart grids, deep learning, and so forth. In this problem,
there are a network of agents who may be cooperative (i.e., decentralized
online optimization) or noncooperative (i.e., online game) through local
information exchanges, and the local cost function of each agent is often
time-varying in dynamic and even adversarial environments. At each time, a
decision must be made by each agent based on historical information at hand
without knowing future information on cost functions. Although this problem has
been extensively studied in the last decade, a comprehensive survey is lacking.
Therefore, this paper provides a thorough overview of DOL from the perspective
of problem settings, communication, computation, and performances. In addition,
some potential future directions are also discussed in details. |
Reward Systems for Trustworthy Medical Federated Learning | Federated learning (FL) has received high interest from researchers and
practitioners to train machine learning (ML) models for healthcare. Ensuring
the trustworthiness of these models is essential. Especially bias, defined as a
disparity in the model's predictive performance across different subgroups, may
cause unfairness against specific subgroups, which is an undesired phenomenon
for trustworthy ML models. In this research, we address the question to which
extent bias occurs in medical FL and how to prevent excessive bias through
reward systems. We first evaluate how to measure the contributions of
institutions toward predictive performance and bias in cross-silo medical FL
with a Shapley value approximation method. In a second step, we design
different reward systems incentivizing contributions toward high predictive
performance or low bias. We then propose a combined reward system that
incentivizes contributions toward both. We evaluate our work using multiple
medical chest X-ray datasets focusing on patient subgroups defined by patient
sex and age. Our results show that we can successfully measure contributions
toward bias, and an integrated reward system successfully incentivizes
contributions toward a well-performing model with low bias. While the
partitioning of scans only slightly influences the overall bias, institutions
with data predominantly from one subgroup introduce a favorable bias for this
subgroup. Our results indicate that reward systems, which focus on predictive
performance only, can transfer model bias against patients to an institutional
level. Our work helps researchers and practitioners design reward systems for
FL with well-aligned incentives for trustworthy ML. |
Training High-Performance Low-Latency Spiking Neural Networks by Differentiation on Spike Representation | Spiking Neural Network (SNN) is a promising energy-efficient AI model when
implemented on neuromorphic hardware. However, it is a challenge to efficiently
train SNNs due to their non-differentiability. Most existing methods either
suffer from high latency (i.e., long simulation time steps), or cannot achieve
as high performance as Artificial Neural Networks (ANNs). In this paper, we
propose the Differentiation on Spike Representation (DSR) method, which could
achieve high performance that is competitive to ANNs yet with low latency.
First, we encode the spike trains into spike representation using (weighted)
firing rate coding. Based on the spike representation, we systematically derive
that the spiking dynamics with common neural models can be represented as some
sub-differentiable mapping. With this viewpoint, our proposed DSR method trains
SNNs through gradients of the mapping and avoids the common
non-differentiability problem in SNN training. Then we analyze the error when
representing the specific mapping with the forward computation of the SNN. To
reduce such error, we propose to train the spike threshold in each layer, and
to introduce a new hyperparameter for the neural models. With these components,
the DSR method can achieve state-of-the-art SNN performance with low latency on
both static and neuromorphic datasets, including CIFAR-10, CIFAR-100, ImageNet,
and DVS-CIFAR10. |
An Analysis of the Features Considerable for NFT Recommendations | This research explores the methods that NFTs can be recommended to people who
interact with NFT-marketplaces to explore NFTs of preference and similarity to
what they have been searching for. While exploring past methods that can be
adopted for recommendations, the use of NFT traits for recommendations has been
explored. The outcome of the research highlights the necessity of using
multiple Recommender Systems to present the user with the best possible NFTs
when interacting with decentralized systems. |
Molecular Identification from AFM images using the IUPAC Nomenclature and Attribute Multimodal Recurrent Neural Networks | Despite being the main tool to visualize molecules at the atomic scale, AFM
with CO-functionalized metal tips is unable to chemically identify the observed
molecules. Here we present a strategy to address this challenging task using
deep learning techniques. Instead of identifying a finite number of molecules
following a traditional classification approach, we define the molecular
identification as an image captioning problem. We design an architecture,
composed of two multimodal recurrent neural networks, capable of identifying
the structure and composition of an unknown molecule using a 3D-AFM image stack
as input. The neural network is trained to provide the name of each molecule
according to the IUPAC nomenclature rules. To train and test this algorithm we
use the novel QUAM-AFM dataset, which contains almost 700,000 molecules and 165
million AFM images. The accuracy of the predictions is remarkable, achieving a
high score quantified by the cumulative BLEU 4-gram, a common metric in
language recognition studies. |
Adaptive Online Optimization with Predictions: Static and Dynamic Environments | In the past few years, Online Convex Optimization (OCO) has received notable
attention in the control literature thanks to its flexible real-time nature and
powerful performance guarantees. In this paper, we propose new step-size rules
and OCO algorithms that simultaneously exploit gradient predictions, function
predictions and dynamics, features particularly pertinent to control
applications. The proposed algorithms enjoy static and dynamic regret bounds in
terms of the dynamics of the reference action sequence, gradient prediction
error and function prediction error, which are generalizations of known
regularity measures from the literature. We present results for both convex and
strongly convex costs. We validate the performance of the proposed algorithms
in a trajectory tracking case study, as well as portfolio optimization using
real-world datasets. |
Differentially Private Multivariate Time Series Forecasting of Aggregated Human Mobility With Deep Learning: Input or Gradient Perturbation? | This paper investigates the problem of forecasting multivariate aggregated
human mobility while preserving the privacy of the individuals concerned.
Differential privacy, a state-of-the-art formal notion, has been used as the
privacy guarantee in two different and independent steps when training deep
learning models. On one hand, we considered \textit{gradient perturbation},
which uses the differentially private stochastic gradient descent algorithm to
guarantee the privacy of each time series sample in the learning stage. On the
other hand, we considered \textit{input perturbation}, which adds differential
privacy guarantees in each sample of the series before applying any learning.
We compared four state-of-the-art recurrent neural networks: Long Short-Term
Memory, Gated Recurrent Unit, and their Bidirectional architectures, i.e.,
Bidirectional-LSTM and Bidirectional-GRU. Extensive experiments were conducted
with a real-world multivariate mobility dataset, which we published openly
along with this paper. As shown in the results, differentially private deep
learning models trained under gradient or input perturbation achieve nearly the
same performance as non-private deep learning models, with loss in performance
varying between $0.57\%$ to $2.8\%$. The contribution of this paper is
significant for those involved in urban planning and decision-making, providing
a solution to the human mobility multivariate forecast problem through
differentially private deep learning models. |
TinyLight: Adaptive Traffic Signal Control on Devices with Extremely Limited Resources | Recent advances in deep reinforcement learning (DRL) have largely promoted
the performance of adaptive traffic signal control (ATSC). Nevertheless,
regarding the implementation, most works are cumbersome in terms of storage and
computation. This hinders their deployment on scenarios where resources are
limited. In this work, we propose TinyLight, the first DRL-based ATSC model
that is designed for devices with extremely limited resources. TinyLight first
constructs a super-graph to associate a rich set of candidate features with a
group of light-weighted network blocks. Then, to diminish the model's resource
consumption, we ablate edges in the super-graph automatically with a novel
entropy-minimized objective function. This enables TinyLight to work on a
standalone microcontroller with merely 2KB RAM and 32KB ROM. We evaluate
TinyLight on multiple road networks with real-world traffic demands.
Experiments show that even with extremely limited resources, TinyLight still
achieves competitive performance. The source code and appendix of this work can
be found at \url{https://bit.ly/38hH8t8}. |
Branch & Learn for Recursively and Iteratively Solvable Problems in Predict+Optimize | This paper proposes Branch & Learn, a framework for Predict+Optimize to
tackle optimization problems containing parameters that are unknown at the time
of solving. Given an optimization problem solvable by a recursive algorithm
satisfying simple conditions, we show how a corresponding learning algorithm
can be constructed directly and methodically from the recursive algorithm. Our
framework applies also to iterative algorithms by viewing them as a degenerate
form of recursion. Extensive experimentation shows better performance for our
proposal over classical and state-of-the-art approaches. |
Uniform Manifold Approximation with Two-phase Optimization | We introduce Uniform Manifold Approximation with Two-phase Optimization
(UMATO), a dimensionality reduction (DR) technique that improves UMAP to
capture the global structure of high-dimensional data more accurately. In
UMATO, optimization is divided into two phases so that the resulting embeddings
can depict the global structure reliably while preserving the local structure
with sufficient accuracy. As the first phase, hub points are identified and
projected to construct a skeletal layout for the global structure. In the
second phase, the remaining points are added to the embedding preserving the
regional characteristics of local areas. Through quantitative experiments, we
found that UMATO (1) outperformed widely used DR techniques in preserving the
global structure while (2) producing competitive accuracy in representing the
local structure. We also verified that UMATO is preferable in terms of
robustness over diverse initialization methods, number of epochs, and
subsampling techniques. |
Don't Blame the Annotator: Bias Already Starts in the Annotation Instructions | In recent years, progress in NLU has been driven by benchmarks. These
benchmarks are typically collected by crowdsourcing, where annotators write
examples based on annotation instructions crafted by dataset creators. In this
work, we hypothesize that annotators pick up on patterns in the crowdsourcing
instructions, which bias them to write similar examples that are then
over-represented in the collected data. We study this form of bias, termed
instruction bias, in 14 recent NLU benchmarks, showing that instruction
examples often exhibit concrete patterns, which are propagated by crowdworkers
to the collected data. This extends previous work (Geva et al., 2019) and
raises a new concern of whether we are modeling the dataset creator's
instructions, rather than the task. Through a series of experiments, we show
that, indeed, instruction bias can lead to overestimation of model performance,
and that models struggle to generalize beyond biases originating in the
crowdsourcing instructions. We further analyze the influence of instruction
bias in terms of pattern frequency and model size, and derive concrete
recommendations for creating future NLU benchmarks. |
DDDM: a Brain-Inspired Framework for Robust Classification | Despite their outstanding performance in a broad spectrum of real-world
tasks, deep artificial neural networks are sensitive to input noises,
particularly adversarial perturbations. On the contrary, human and animal
brains are much less vulnerable. In contrast to the one-shot inference
performed by most deep neural networks, the brain often solves decision-making
with an evidence accumulation mechanism that may trade time for accuracy when
facing noisy inputs. The mechanism is well described by the Drift-Diffusion
Model (DDM). In the DDM, decision-making is modeled as a process in which noisy
evidence is accumulated toward a threshold. Drawing inspiration from the DDM,
we propose the Dropout-based Drift-Diffusion Model (DDDM) that combines
test-phase dropout and the DDM for improving the robustness for arbitrary
neural networks. The dropouts create temporally uncorrelated noises in the
network that counter perturbations, while the evidence accumulation mechanism
guarantees a reasonable decision accuracy. Neural networks enhanced with the
DDDM tested in image, speech, and text classification tasks all significantly
outperform their native counterparts, demonstrating the DDDM as a task-agnostic
defense against adversarial attacks. |
A Simple Approach to Improve Single-Model Deep Uncertainty via Distance-Awareness | Accurate uncertainty quantification is a major challenge in deep learning, as
neural networks can make overconfident errors and assign high confidence
predictions to out-of-distribution (OOD) inputs. The most popular approaches to
estimate predictive uncertainty in deep learning are methods that combine
predictions from multiple neural networks, such as Bayesian neural networks
(BNNs) and deep ensembles. However their practicality in real-time,
industrial-scale applications are limited due to the high memory and
computational cost. Furthermore, ensembles and BNNs do not necessarily fix all
the issues with the underlying member networks. In this work, we study
principled approaches to improve uncertainty property of a single network,
based on a single, deterministic representation. By formalizing the uncertainty
quantification as a minimax learning problem, we first identify distance
awareness, i.e., the model's ability to quantify the distance of a testing
example from the training data, as a necessary condition for a DNN to achieve
high-quality (i.e., minimax optimal) uncertainty estimation. We then propose
Spectral-normalized Neural Gaussian Process (SNGP), a simple method that
improves the distance-awareness ability of modern DNNs with two simple changes:
(1) applying spectral normalization to hidden weights to enforce bi-Lipschitz
smoothness in representations and (2) replacing the last output layer with a
Gaussian process layer. On a suite of vision and language understanding
benchmarks, SNGP outperforms other single-model approaches in prediction,
calibration and out-of-domain detection. Furthermore, SNGP provides
complementary benefits to popular techniques such as deep ensembles and data
augmentation, making it a simple and scalable building block for probabilistic
deep learning. Code is open-sourced at
https://github.com/google/uncertainty-baselines |
Fine-Grained Address Segmentation for Attention-Based Variable-Degree Prefetching | Machine learning algorithms have shown potential to improve prefetching
performance by accurately predicting future memory accesses. Existing
approaches are based on the modeling of text prediction, considering
prefetching as a classification problem for sequence prediction. However, the
vast and sparse memory address space leads to large vocabulary, which makes
this modeling impractical. The number and order of outputs for multiple cache
line prefetching are also fundamentally different from text prediction. We
propose TransFetch, a novel way to model prefetching. To reduce vocabulary
size, we use fine-grained address segmentation as input. To predict unordered
sets of future addresses, we use delta bitmaps for multiple outputs. We apply
an attention-based network to learn the mapping between input and output.
Prediction experiments demonstrate that address segmentation achieves 26% - 36%
higher F1-score than delta inputs and 15% - 24% higher F1-score than page &
offset inputs for SPEC 2006, SPEC 2017, and GAP benchmarks. Simulation results
show that TransFetch achieves 38.75% IPC improvement compared with no
prefetching, outperforming the best-performing rule-based prefetcher BOP by
10.44%, and ML-based prefetcher Voyager by 6.64%. |
A Word is Worth A Thousand Dollars: Adversarial Attack on Tweets Fools Stock Prediction | More and more investors and machine learning models rely on social media
(e.g., Twitter and Reddit) to gather real-time information and sentiment to
predict stock price movements. Although text-based models are known to be
vulnerable to adversarial attacks, whether stock prediction models have similar
vulnerability is underexplored. In this paper, we experiment with a variety of
adversarial attack configurations to fool three stock prediction victim models.
We address the task of adversarial generation by solving combinatorial
optimization problems with semantics and budget constraints. Our results show
that the proposed attack method can achieve consistent success rates and cause
significant monetary loss in trading simulation by simply concatenating a
perturbed but semantically similar tweet. |
Processing Network Controls via Deep Reinforcement Learning | Novel advanced policy gradient (APG) algorithms, such as proximal policy
optimization (PPO), trust region policy optimization, and their variations,
have become the dominant reinforcement learning (RL) algorithms because of
their ease of implementation and good practical performance. This dissertation
is concerned with theoretical justification and practical application of the
APG algorithms for solving processing network control optimization problems.
Processing network control problems are typically formulated as Markov decision
process (MDP) or semi-Markov decision process (SMDP) problems that have several
unconventional for RL features: infinite state spaces, unbounded costs,
long-run average cost objectives. Policy improvement bounds play a crucial role
in the theoretical justification of the APG algorithms. In this thesis we
refine existing bounds for MDPs with finite state spaces and prove novel policy
improvement bounds for classes of MDPs and SMDPs used to model processing
network operations. We consider two examples of processing network control
problems and customize the PPO algorithm to solve them. First, we consider
parallel-server and multiclass queueing networks controls. Second, we consider
the drivers repositioning problem in a ride-hailing service system. For both
examples the PPO algorithm with auxiliary modifications consistently generates
control policies that outperform state-of-art heuristics. |
Neural Network Optimal Feedback Control with Guaranteed Local Stability | Recent research shows that deep learning can be an effective tool for
designing optimal feedback controllers for high-dimensional nonlinear dynamic
systems. But the behavior of these neural network (NN) controllers is still not
well understood. In particular, some NNs with high test accuracy can fail to
even locally stabilize the dynamic system. To address this challenge we propose
several novel NN architectures, which we show guarantee local stability while
retaining the semi-global approximation capacity to learn the optimal feedback
policy. The proposed architectures are compared against standard NN feedback
controllers through numerical simulations of two high-dimensional nonlinear
optimal control problems (OCPs): stabilization of an unstable Burgers-type
partial differential equation (PDE), and altitude and course tracking for a six
degree-of-freedom (6DoF) unmanned aerial vehicle (UAV). The simulations
demonstrate that standard NNs can fail to stabilize the dynamics even when
trained well, while the proposed architectures are always at least locally
stable. Moreover, the proposed controllers are found to be near-optimal in
testing. |
Abnormal-aware Multi-person Evaluation System with Improved Fuzzy Weighting | There exists a phenomenon that subjectivity highly lies in the daily
evaluation process. Our research primarily concentrates on a multi-person
evaluation system with anomaly detection to minimize the possible inaccuracy
that subjective assessment brings. We choose the two-stage screening method,
which consists of rough screening and score-weighted Kendall-$\tau$ Distance to
winnow out abnormal data, coupled with hypothesis testing to narrow global
discrepancy. Then we use Fuzzy Synthetic Evaluation Method(FSE) to determine
the significance of scores given by reviewers as well as their reliability,
culminating in a more impartial weight for each reviewer in the final
conclusion. The results demonstrate a clear and comprehensive ranking instead
of unilateral scores, and we get to have an efficiency in filtering out
abnormal data as well as a reasonably objective weight determination mechanism.
We can sense that through our study, people will have a chance of modifying a
multi-person evaluation system to attain both equity and a relatively superior
competitive atmosphere. |
Detecting COVID-19 Conspiracy Theories with Transformers and TF-IDF | The sharing of fake news and conspiracy theories on social media has
wide-spread negative effects. By designing and applying different machine
learning models, researchers have made progress in detecting fake news from
text. However, existing research places a heavy emphasis on general,
common-sense fake news, while in reality fake news often involves rapidly
changing topics and domain-specific vocabulary. In this paper, we present our
methods and results for three fake news detection tasks at MediaEval benchmark
2021 that specifically involve COVID-19 related topics. We experiment with a
group of text-based models including Support Vector Machines, Random Forest,
BERT, and RoBERTa. We find that a pre-trained transformer yields the best
validation results, but a randomly initialized transformer with smart design
can also be trained to reach accuracies close to that of the pre-trained
transformer. |
Fair Feature Subset Selection using Multiobjective Genetic Algorithm | The feature subset selection problem aims at selecting the relevant subset of
features to improve the performance of a Machine Learning (ML) algorithm on
training data. Some features in data can be inherently noisy, costly to
compute, improperly scaled, or correlated to other features, and they can
adversely affect the accuracy, cost, and complexity of the induced algorithm.
The goal of traditional feature selection approaches has been to remove such
irrelevant features. In recent years ML is making a noticeable impact on the
decision-making processes of our everyday lives. We want to ensure that these
decisions do not reflect biased behavior towards certain groups or individuals
based on protected attributes such as age, sex, or race. In this paper, we
present a feature subset selection approach that improves both fairness and
accuracy objectives and computes Pareto-optimal solutions using the NSGA-II
algorithm. We use statistical disparity as a fairness metric and F1-Score as a
metric for model performance. Our experiments on the most commonly used
fairness benchmark datasets with three different machine learning algorithms
show that using the evolutionary algorithm we can effectively explore the
trade-off between fairness and accuracy. |
A Simple Duality Proof for Wasserstein Distributionally Robust Optimization | We present a short and elementary proof of the duality for Wasserstein
distributionally robust optimization, which holds for any arbitrary Kantorovich
transport distance, any arbitrary measurable loss function, and any arbitrary
nominal probability distribution, as long as certain interchangeability
principle holds. |
Combined Learning of Neural Network Weights for Privacy in Collaborative Tasks | We introduce CoLN, Combined Learning of Neural network weights, a novel
method to securely combine Machine Learning models over sensitive data with no
sharing of data. With CoLN, local hosts use the same Neural Network
architecture and base parameters to train a model using only locally available
data. Locally trained models are then submitted to a combining agent, which
produces a combined model. The new model's parameters can be sent back to
hosts, and can then be used as initial parameters for a new training iteration.
CoLN is capable of combining several distributed neural networks of the same
kind but is not restricted to any single neural architecture. In this paper we
detail the combination algorithm and present experiments with feed-forward,
convolutional, and recurrent Neural Network architectures, showing that the
CoLN combined model approximates the performance of a hypothetical ideal
centralized model, trained using the combination of the local datasets. CoLN
can contribute to secure collaborative research, as required in the medical
area, where privacy issues preclude data sharing, but where the limitations of
local data demand information derived from larger datasets. |
Adapting and Evaluating Influence-Estimation Methods for Gradient-Boosted Decision Trees | Influence estimation analyzes how changes to the training data can lead to
different model predictions; this analysis can help us better understand these
predictions, the models making those predictions, and the data sets they're
trained on. However, most influence-estimation techniques are designed for deep
learning models with continuous parameters. Gradient-boosted decision trees
(GBDTs) are a powerful and widely-used class of models; however, these models
are black boxes with opaque decision-making processes. In the pursuit of better
understanding GBDT predictions and generally improving these models, we adapt
recent and popular influence-estimation methods designed for deep learning
models to GBDTs. Specifically, we adapt representer-point methods and TracIn,
denoting our new methods TREX and BoostIn, respectively; source code is
available at https://github.com/jjbrophy47/tree_influence. We compare these
methods to LeafInfluence and other baselines using 5 different evaluation
measures on 22 real-world data sets with 4 popular GBDT implementations. These
experiments give us a comprehensive overview of how different approaches to
influence estimation work in GBDT models. We find BoostIn is an efficient
influence-estimation method for GBDTs that performs equally well or better than
existing work while being four orders of magnitude faster. Our evaluation also
suggests the gold-standard approach of leave-one-out~(LOO) retraining
consistently identifies the single-most influential training example but
performs poorly at finding the most influential set of training examples for a
given target prediction. |