Publications

Towards Causal Federated Learning For Enhanced Robustness and Privacy
Sreya Francis
Irene Tenison
Irina Rish
Federated Learning is an emerging privacy-preserving distributed machine learning approach to building a shared model by performing distribu… (voir plus)ted training locally on participating devices (clients) and aggregating the local models into a global one. As this approach prevents data collection and aggregation, it helps in reducing associated privacy risks to a great extent. However, the data samples across all participating clients are usually not independent and identically distributed (non-iid), and Out of Distribution(OOD) generalization for the learned models can be poor. Besides this challenge, federated learning also remains vulnerable to various attacks on security wherein a few malicious participating entities work towards inserting backdoors, degrading the generated aggregated model as well as inferring the data owned by participating entities. In this paper, we propose an approach for learning invariant (causal) features common to all participating clients in a federated learning setup and analyze empirically how it enhances the Out of Distribution (OOD) accuracy as well as the privacy of the final learned model.
2021-04-13
ArXiv (prépublication)
arxiv.org
arxiv.org
An Analysis of the Adaptation Speed of Causal Models
Rémi Le Priol
Reza Babanezhad Harikandeh
Yoshua Bengio
Simon Lacoste-Julien
Consider a collection of datasets generated by unknown interventions on an unknown structural causal model …
2021-04-12
International Conference on Artificial Intelligence and Statistics (Poster)
doi.org
proceedings.mlr.press
Neural Function Modules with Sparse Arguments: A Dynamic Approach to Integrating Information across Layers
Alex Lamb
Anirudh Goyal
Agnieszka Słowik
Michael Mozer
Philippe Beaudoin
Yoshua Bengio
Feed-forward neural networks consist of a sequence of layers, in which each layer performs some processing on the information from the previ… (voir plus)ous layer. A downside to this approach is that each layer (or module, as multiple modules can operate in parallel) is tasked with processing the entire hidden state, rather than a particular part of the state which is most relevant for that module. Methods which only operate on a small number of input variables are an essential part of most programming languages, and they allow for improved modularity and code re-usability. Our proposed method, Neural Function Modules (NFM), aims to introduce the same structural capability into deep learning. Most of the work in the context of feed-forward networks combining top-down and bottom-up feedback is limited to classification problems. The key contribution of our work is to combine attention, sparsity, top-down and bottom-up feedback, in a flexible algorithm which, as we show, improves the results in standard classification, out-of-domain generalization, generative modeling, and learning representations in the context of reinforcement learning.
2021-04-12
International Conference on Artificial Intelligence and Statistics (Poster)
doi.org
proceedings.mlr.press
Science-Software Linkage: The Challenges of Traceability between Scientific Knowledge and Software Artifacts
Hideaki Hata
Jin L.C. Guo
Raula Gaikovina Kula
Christoph Treude
Although computer science papers are often accompanied by software artifacts, connecting research papers to their software artifacts and vic… (voir plus)e versa is not always trivial. First of all, there is a lack of well-accepted standards for how such links should be provided. Furthermore, the provided links, if any, often become outdated: they are affected by link rot when pre-prints are removed, when repositories are migrated, or when papers and repositories evolve independently. In this paper, we summarize the state of the practice of linking research papers and associated source code, highlighting the recent efforts towards creating and maintaining such links. We also report on the results of several empirical studies focusing on the relationship between scientific papers and associated software artifacts, and we outline challenges related to traceability and opportunities for overcoming these challenges.
2021-04-12
ArXiv (prépublication)
arxiv.org
arxiv.org
Common Limitations of Image Processing Metrics: A Picture Story
Annika Reinke
Matthias Eisenmann
Minu Dietlinde Tizabi
Carole H. Sudre
TIM RÄDSCH
Michela Antonelli
Tal Arbel
Spyridon Bakas
M. Jorge Cardoso
Veronika Cheplygina
Keyvan Farahani
B. Glocker
DOREEN HECKMANN-NÖTZEL
Fabian Isensee
Pierre Jannin
Charles E. Jr. Kahn
Jens Kleesiek
Tahsin Kurc
Michal Kozubek
Bennett Landman … (voir 14 de plus)
GEERT LITJENS
Klaus Maier-Hein
Bjoern Menze
Henning Müller
Jens Petersen
Mauricio Reyes
Nicola Rieke
Bram Stieltjes
R. Summers
Sotirios A. Tsaftaris
Bram van Ginneken
Annette Kopp-Schneider
PAUL F. JÄGER
Lena Maier-Hein
2021-04-11
ArXiv (prépublication)
arxiv.org
arxiv.org
Maintenance of a collection of machines under partial observability: Indexability and computation of Whittle index
Nima Akbarzadeh
Aditya Mahajan
We consider the problem of scheduling maintenance for a collection of machines under partial observations when the state of each machine det… (voir plus)eriorates stochastically in a Markovian manner. We consider two observational models: first, the state of each machine is not observable at all, and second, the state of each machine is observable only if a service-person visits them. The agent takes a maintenance action, e.g., machine replacement, if he is chosen for the task. We model both problems as restless multi-armed bandit problem and propose the Whittle index policy for scheduling the visits. We show that both models are indexable. For the first model, we derive a closed-form expression for the Whittle index. For the second model, we propose an efficient algorithm to compute the Whittle index by exploiting the qualitative properties of the optimal policy. We present detailed numerical experiments which show that for multiple instances of the model, the Whittle index policy outperforms myopic policy and can be close-to-optimal in different setups.
2021-04-11
arXiv.org (prépublication)
dblp.uni-trier.de
Models of Human Behavioral Agents in Bandits, Contextual Bandits and RL
Baihan Lin
Guillermo Cecchi
Djallel Bouneffouf
Jenna Reinen
Irina Rish
2021-04-07
Human Brain and Artificial Intelligence (publié)
doi.org
Safe Option-Critic: Learning Safety in the Option-Critic Architecture
Arushi Jain
Khimya Khetarpal
Doina Precup
Designing hierarchical reinforcement learning algorithms that exhibit safe behaviour is not only vital for practical applications but also, … (voir plus)facilitates a better understanding of an agent's decisions. We tackle this problem in the options framework, a particular way to specify temporally abstract actions which allow an agent to use sub-policies with start and end conditions. We consider a behaviour as safe that avoids regions of state-space with high uncertainty in the outcomes of actions. We propose an optimization objective that learns safe options by encouraging the agent to visit states with higher behavioural consistency. The proposed objective results in a trade-off between maximizing the standard expected return and minimizing the effect of model uncertainty in the return. We propose a policy gradient algorithm to optimize the constrained objective function. We examine the quantitative and qualitative behaviour of the proposed approach in a tabular grid-world, continuous-state puddle-world, and three games from the Arcade Learning Environment: Ms.Pacman, Amidar, and Q*Bert. Our approach achieves a reduction in the variance of return, boosts performance in environments with intrinsic variability in the reward structure, and compares favorably both with primitive actions as well as with risk-neutral options.
2021-04-06
The Knowledge Engineering Review (publié)
doi.org
arxiv.org
Accounting for Variance in Machine Learning Benchmarks
Xavier Bouthillier
Pierre Delaunay
Mirko Bronzi
Assya Trofimov
Brennan Nichyporuk
Justin Szeto
Naz Sepah
Edward Raff
Kanika Madan
Vikram Voleti
Samira Ebrahimi Kahou
Vincent Michalski
Dmitriy Serdyuk
Tal Arbel
Chris Pal
Gael Varoquaux
Pascal Vincent
Strong empirical evidence that one machine-learning algorithm A outperforms another one B ideally calls for multiple trials optimizing the l… (voir plus)earning pipeline over sources of variation such as data sampling, data augmentation, parameter initialization, and hyperparameters choices. This is prohibitively expensive, and corners are cut to reach conclusions. We model the whole benchmarking process, revealing that variance due to data sampling, parameter initialization and hyperparameter choice impact markedly the results. We analyze the predominant comparison methods used today in the light of this variance. We show a counter-intuitive result that adding more sources of variation to an imperfect estimator approaches better the ideal estimator at a 51 times reduction in compute cost. Building on these results, we study the error rate of detecting improvements, on five different deep-learning tasks/architectures. This study leads us to propose recommendations for performance comparisons.
2021-04-05
Conference on Machine Learning and Systems (Accept)
doi.org
arxiv.org
Comparing Transfer and Meta Learning Approaches on a Unified Few-Shot Classification Benchmark
Vincent Dumoulin
Neil Houlsby
Utku Evci
Xiaohua Zhai
Ross Goroshin
Sylvain Gelly
Hugo Larochelle
Meta and transfer learning are two successful families of approaches to few-shot learning. Despite highly related goals, state-of-the-art ad… (voir plus)vances in each family are measured largely in isolation of each other. As a result of diverging evaluation norms, a direct or thorough comparison of different approaches is challenging. To bridge this gap, we perform a cross-family study of the best transfer and meta learners on both a large-scale meta-learning benchmark (Meta-Dataset, MD), and a transfer learning benchmark (Visual Task Adaptation Benchmark, VTAB). We find that, on average, large-scale transfer methods (Big Transfer, BiT) outperform competing approaches on MD, even when trained only on ImageNet. In contrast, meta-learning approaches struggle to compete on VTAB when trained and validated on MD. However, BiT is not without limitations, and pushing for scale does not improve performance on highly out-of-distribution MD tasks. In performing this study, we reveal a number of discrepancies in evaluation norms and study some of these in light of the performance gap. We hope that this work facilitates sharing of insights from each community, and accelerates progress on few-shot learning.
2021-04-05
ArXiv (prépublication)
arxiv.org
arxiv.org
Understanding Continual Learning Settings with Data Distribution Drift Analysis
Timothee LESORT
Massimo Caccia
Irina Rish
Classical machine learning algorithms often assume that the data are drawn i.i.d. from a stationary probability distribution. Recently, cont… (voir plus)inual learning emerged as a rapidly growing area of machine learning where this assumption is relaxed, i.e. where the data distribution is non-stationary and changes over time. This paper represents the state of data distribution by a context variable
2021-04-03
ArXiv (prépublication)
arxiv.org
arxiv.org
Differentially private data publishing for arbitrarily partitioned data
Rongli Wang
Benjamin C. M. Fung
Yan Zhu
Qiang Peng
2021-03-31
Information Sciences (publié)
doi.org