Publications

Asymmetric Actor-Critic with Approximate Information State
Amit Sinha
Aditya Mahajan
Reinforcement learning (RL) for partially observable Markov decision processes (POMDPs) is a challenging problem because decisions need to b… (voir plus)e made based on the entire history of observations and actions. However, in several scenarios, state information is available during the training phase. We are interested in exploiting the availability of this state information during the training phase to efficiently learn a history-based policy using RL. Specifically, we consider actor-critic algorithms, where the actor uses only the history information but the critic uses both history and state. Such algorithms are called asymmetric actor-critic, to highlight the fact that the actor and critic have asymmetric information. Motivated by the recent success of using representation losses in RL for POMDPs [1], we derive similar theoretical results for the asymmetric actor-critic case and evaluate the effectiveness of adding such auxiliary losses in experiments. In particular, we learn a history representation-called an approximate information state (AIS)-and bound the performance loss when acting using AIS.
2023-12-12
IEEE Conference on Decision and Control (publié)
doi.org
Current Practices in Voice Data Collection and Limitations to Voice AI Research: A National Survey.
Emily Evangelista
Rohan Kale
Desiree McCutcheon
Anais Rameau
Alexander Gelbard
Maria Powell
Michael Johns
Anthony Law
Phillip Song
Matthew Naunheim
Stephanie Watts
Paul C. Bryson
Matthew G. Crowson
Jeremy Pinto
Yael Bensoussan
INTRODUCTION Accuracy and validity of voice AI algorithms rely on substantial quality voice data. Although commensurable amounts of voice da… (voir plus)ta are captured daily in voice centers across North America, there is no standardized protocol for acoustic data management, which limits the usability of these datasets for voice artificial intelligence (AI) research. OBJECTIVE The aim was to capture current practices of voice data collection, storage, analysis, and perceived limitations to collaborative voice research. METHODS A 30-question online survey was developed with expert guidance from the voicecollab.ai members, an international collaborative of voice AI researchers. The survey was disseminated via REDCap to an estimated 200 practitioners at North American voice centers. Survey questions assessed respondents' current practices in terms of acoustic data collection, storage, and retrieval as well as limitations to collaborative voice research. RESULTS Seventy-two respondents completed the survey of which 81.7% were laryngologists and 18.3% were speech language pathologists (SLPs). Eighteen percent of respondents reported seeing 40%-60% and 55% reported seeing >60 patients with voice disorders weekly (conservative estimate of over 4000 patients/week). Only 28% of respondents reported utilizing standardized protocols for collection and storage of acoustic data. Although, 87% of respondents conduct voice research, only 38% of respondents report doing so on a multi-institutional level. Perceived limitations to conducting collaborative voice research include lack of standardized methodology for collection (30%) and lack of human resources to prepare and label voice data adequately (55%). CONCLUSION To conduct large-scale multi-institutional voice research with AI, there is a pertinent need for standardization of acoustic data management, as well as an infrastructure for secure and efficient data sharing. LEVEL OF EVIDENCE Level 5 Laryngoscope, 2023.
2023-12-12
The Laryngoscope (publié)
doi.org
Detection and evaluation of bias-inducing features in machine learning
Moses Openja
Gabriel Laberge
Foutse Khomh
2023-12-12
Empirical Software Engineering (publié)
doi.org
arxiv.org
Preserving Privacy in GANs Against Membership Inference Attack
Mohammadhadi Shateri
Francisco Messina
Fabrice Labeau
Pablo Piantanida
Generative Adversarial Networks (GANs) have been widely used for generating synthetic data for cases where there is a limited size real-worl… (voir plus)d dataset or when data holders are unwilling to share their data samples. Recent works showed that GANs, due to overfitting and memorization, might leak information regarding their training data samples. This makes GANs vulnerable to Membership Inference Attacks (MIAs). Several defense strategies have been proposed in the literature to mitigate this privacy issue. Unfortunately, defense strategies based on differential privacy are proven to reduce extensively the quality of the synthetic data points. On the other hand, more recent frameworks such as PrivGAN and PAR-GAN are not suitable for small-size training datasets. In the present work, the overfitting in GANs is studied in terms of the discriminator, and a more general measure of overfitting based on the Bhattacharyya coefficient is defined. Then, inspired by Fano's inequality, our first defense mechanism against MIAs is proposed. This framework, which requires only a simple modification in the loss function of GANs, is referred to as the maximum entropy GAN or MEGAN and significantly improves the robustness of GANs to MIAs. As a second defense strategy, a more heuristic model based on minimizing the information leaked from generated samples about the training data points is presented. This approach is referred to as mutual information minimization GAN (MIMGAN) and uses a variational representation of the mutual information to minimize the information that a synthetic sample might leak about the whole training data set. Applying the proposed frameworks to some commonly used data sets against state-of-the-art MIAs reveals that the proposed methods can reduce the accuracy of the adversaries to the level of random guessing accuracy with a small reduction in the quality of the synthetic data samples.
2023-12-12
IEEE Transactions on Information Forensics and Security (inconnu)
doi.org
arxiv.org
Privacy-preserving analysis of time-to-event data under nested case-control sampling
Lamin Juwara
Archer Yang
Ana M Velly
Paramita Saha-Chaudhuri
2023-12-12
Statistical Methods in Medical Research (publié)
doi.org
Relative Almost Sure Regret Bounds for Certainty Equivalence Control of Markov Jump Systems
Borna Sayedana
Mohammad Afshari
Peter E. Caines
Aditya Mahajan
In this paper, we consider learning and control problem in an unknown Markov jump linear system (MJLS) with perfect state observations. We f… (voir plus)irst establish a generic upper bound on regret for any learning based algorithm. We then propose a certainty equivalence-based learning alagrithm and show that this algorithm achieves a regret of
2023-12-12
2023 62nd IEEE Conference on Decision and Control (CDC) (publié)
doi.org
Weighted-Norm Bounds on Model Approximation in MDPs with Unbounded Per-Step Cost
Berk Bozkurt
Aditya Mahajan
Ashutosh Nayyar
Yi Ouyang
We consider the problem of designing a control policy for an infinite-horizon discounted cost Markov Decision Process (MDP) …
2023-12-12
IEEE Conference on Decision and Control (publié)
doi.org
A²CiD²: Accelerating Asynchronous Communication in Decentralized Deep Learning
Adel Nabli
Eugene Belilovsky
Edouard Oyallon
Distributed training of Deep Learning models has been critical to many recent successes in the field. Current standard methods primarily rel… (voir plus)y on synchronous centralized algorithms which induce major communication bottlenecks and synchronization locks at scale. Decentralized asynchronous algorithms are emerging as a potential alternative but their practical applicability still lags. In order to mitigate the increase in communication cost that naturally comes with scaling the number of workers, we introduce a principled asynchronous, randomized, gossip-based optimization algorithm which works thanks to a continuous local momentum named
2023-12-11
Conference on Neural Information Processing Systems (Accept (poster))
doi.org
openreview.net
Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolation
Sébastien Lachapelle
Divyat Mahajan
Ioannis Mitliagkas
Simon Lacoste-Julien
We tackle the problems of latent variables identification and ``out-of-support'' image generation in representation learning. We show that b… (voir plus)oth are possible for a class of decoders that we call additive, which are reminiscent of decoders used for object-centric representation learning (OCRL) and well suited for images that can be decomposed as a sum of object-specific images. We provide conditions under which exactly solving the reconstruction problem using an additive decoder is guaranteed to identify the blocks of latent variables up to permutation and block-wise invertible transformations. This guarantee relies only on very weak assumptions about the distribution of the latent factors, which might present statistical dependencies and have an almost arbitrarily shaped support. Our result provides a new setting where nonlinear independent component analysis (ICA) is possible and adds to our theoretical understanding of OCRL methods. We also show theoretically that additive decoders can generate novel images by recombining observed factors of variations in novel ways, an ability we refer to as Cartesian-product extrapolation. We show empirically that additivity is crucial for both identifiability and extrapolation on simulated data.
2023-12-11
Neural Information Processing Systems (Accept (oral))
doi.org
openreview.net
A*Net: A Scalable Path-based Reasoning Approach for Knowledge Graphs
Zhaocheng Zhu
Xinyu Yuan
Mikhail Galkin
Sophie Xhonneux
Ming Zhang
Maxime Gazeau
Jian Tang
Reasoning on large-scale knowledge graphs has been long dominated by embedding methods. While path-based methods possess the inductive capac… (voir plus)ity that embeddings lack, their scalability is limited by the exponential number of paths. Here we present A*Net, a scalable path-based method for knowledge graph reasoning. Inspired by the A* algorithm for shortest path problems, our A*Net learns a priority function to select important nodes and edges at each iteration, to reduce time and memory footprint for both training and inference. The ratio of selected nodes and edges can be specified to trade off between performance and efficiency. Experiments on both transductive and inductive knowledge graph reasoning benchmarks show that A*Net achieves competitive performance with existing state-of-the-art path-based methods, while merely visiting 10% nodes and 10% edges at each iteration. On a million-scale dataset ogbl-wikikg2, A*Net not only achieves a new state-of-the-art result, but also converges faster than embedding methods. A*Net is the first path-based method for knowledge graph reasoning at such scale.
2023-12-11
Neural Information Processing Systems (Accept (poster))
doi.org
openreview.net
Are Diffusion Models Vision-And-Language Reasoners?
Benno Krojer
Elinor Poole-Dayan
Vikram Voleti
Christopher Pal
Siva Reddy
Text-conditioned image generation models have recently shown immense qualitative success using denoising diffusion processes. However, unlik… (voir plus)e discriminative vision-and-language models, it is a non-trivial task to subject these diffusion-based generative models to automatic fine-grained quantitative evaluation of high-level phenomena such as compositionality. Towards this goal, we perform two innovations. First, we transform diffusion-based models (in our case, Stable Diffusion) for any image-text matching (ITM) task using a novel method called DiffusionITM. Second, we introduce the Generative-Discriminative Evaluation Benchmark (GDBench) benchmark with 7 complex vision-and-language tasks, bias evaluation and detailed analysis. We find that Stable Diffusion + DiffusionITM is competitive on many tasks and outperforms CLIP on compositional tasks like like CLEVR and Winoground. We further boost its compositional performance with a transfer setup by fine-tuning on MS-COCO while retaining generative capabilities. We also measure the stereotypical bias in diffusion models, and find that Stable Diffusion 2.1 is, for the most part, less biased than Stable Diffusion 1.5. Overall, our results point in an exciting direction bringing discriminative and generative model evaluation closer. We will release code and benchmark setup soon.
2023-12-11
Neural Information Processing Systems (Accept (poster))
doi.org
openreview.net
Block-State Transformers
Jonathan Pilault
Mahan Fathi
Orhan Firat
Christopher Pal
Pierre-Luc Bacon
Ross Goroshin
State space models (SSMs) have shown impressive results on tasks that require modeling long-range dependencies and efficiently scale to long… (voir plus) sequences owing to their subquadratic runtime complexity. Originally designed for continuous signals, SSMs have shown superior performance on a plethora of tasks, in vision and audio; however, SSMs still lag Transformer performance in Language Modeling tasks. In this work, we propose a hybrid layer named Block-State Transformer (BST), that internally combines an SSM sublayer for long-range contextualization, and a Block Transformer sublayer for short-term representation of sequences. We study three different, and completely parallelizable, variants that integrate SSMs and block-wise attention. We show that our model outperforms similar Transformer-based architectures on language modeling perplexity and generalizes to longer sequences. In addition, the Block-State Transformer demonstrates more than tenfold increase in speed at the layer level compared to the Block-Recurrent Transformer when model parallelization is employed.
2023-12-11
Neural Information Processing Systems (Accept (poster))
doi.org
openreview.net