Publications

Knowledge Distillation for Federated Learning: a Practical Guide
Alessio Mora
Irene Tenison
Paolo Bellavista
Irina Rish
Federated Learning (FL) enables the training of Deep Learning models without centrally collecting possibly sensitive raw data. This paves th… (see more)e way for stronger privacy guarantees when building predictive models. The most used algorithms for FL are parameter-averaging based schemes (e.g., Federated Averaging) that, however, have well known limits: (i) Clients must implement the same model architecture; (ii) Transmitting model weights and model updates implies high communication cost, which scales up with the number of model parameters; (iii) In presence of non-IID data distributions, parameter-averaging aggregation schemes perform poorly due to client model drifts. Federated adaptations of regular Knowledge Distillation (KD) can solve and/or mitigate the weaknesses of parameter-averaging FL algorithms while possibly introducing other trade-offs. In this article, we provide a review of KD-based algorithms tailored for specific FL issues.
2022-11-09
ArXiv (preprint)
doi.org
arxiv.org
A debriefing tool to acquire non-technical skills in trauma courses
Fabio Botelho
Jason M. Harley
Natalie Yanchar
Simone Abib
Ilana Bank
Dan Poenaru
2022-11-07
Surgery Open Science (published)
doi.org
Multi-Head Adapter Routing for Cross-Task Generalization
Lucas Caccia
Edoardo Ponti
Zhan Su
Matheus Pereira
Nicolas Le Roux
Alessandro Sordoni
Parameter-efficient fine-tuning (PEFT) for cross-task generalization consists in pre-training adapters on a multi-task training set before f… (see more)ew-shot adaptation to test tasks. Polytropon [Ponti et al., 2023] (
2022-11-07
ArXiv (preprint)
arxiv.org
arxiv.org
PaReco: patched clones and missed patches among the divergent variants of a software family
Poedjadevie Kadjel Ramkisoen
John Businge
Brent van Bladel
Alexandre Decan
Serge Demeyer
Coen De Roover
Foutse Khomh
Re-using whole repositories as a starting point for new projects is often done by maintaining a variant fork parallel to the original. Howev… (see more)er, the common artifacts between both are not always kept up to date. As a result, patches are not optimally integrated across the two repositories, which may lead to sub-optimal maintenance between the variant and the original project. A bug existing in both repositories can be patched in one but not the other (we see this as a missed opportunity) or it can be manually patched in both probably by different developers (we see this as effort duplication). In this paper we present a tool (named PaReCo) which relies on clone detection to mine cases of missed opportunity and effort duplication from a pool of patches. We analyzed 364 (source to target) variant pairs with 8,323 patches resulting in a curated dataset containing 1,116 cases of effort duplication and 1,008 cases of missed opportunities. We achieve a precision of 91%, recall of 80%, accuracy of 88%, and F1-score of 85%. Furthermore, we investigated the time interval between patches and found out that, on average, missed patches in the target variants have been introduced in the source variants 52 weeks earlier. Consequently, PaReCo can be used to manage variability in “time” by automatically identifying interesting patches in later project releases to be backported to supported earlier releases.
2022-11-07
ESEC/SIGSOFT FSE (published)
doi.org
Bayesian learning of Causal Structure and Mechanisms with GFlowNets and Variational Bayes
Mizu Nishikawa-Toomey
Tristan Deleu
Jithendaraa Subramanian
Yoshua Bengio
Laurent Charlin
Bayesian causal structure learning aims to learn a posterior distribution over directed acyclic graphs (DAGs), and the mechanisms that defin… (see more)e the relationship between parent and child variables. By taking a Bayesian approach, it is possible to reason about the uncertainty of the causal model. The notion of modelling the uncertainty over models is particularly crucial for causal structure learning since the model could be unidentifiable when given only a finite amount of observational data. In this paper, we introduce a novel method to jointly learn the structure and mechanisms of the causal model using Variational Bayes, which we call Variational Bayes-DAG-GFlowNet (VBG). We extend the method of Bayesian causal structure learning using GFlowNets to learn not only the posterior distribution over the structure, but also the parameters of a linear-Gaussian model. Our results on simulated data suggest that VBG is competitive against several baselines in modelling the posterior over DAGs and mechanisms, while offering several advantages over existing methods, including the guarantee to sample acyclic graphs, and the flexibility to generalize to non-linear causal mechanisms.
2022-11-04
ArXiv (preprint)
doi.org
arxiv.org
Existing eHealth Solutions for Older Adults Living With Neurocognitive Disorders (Mild and Major) or Dementia and Their Informal Caregivers: Protocol for an Environmental Scan
Ambily Jose
Maxime Sasseville
Samantha Dequanter
Ellen Gorus
Anik Giguère
Anne Bourbonnais
Samira Abbasgholizadeh-Rahimi
Ronald Buyl
Marie-Pierre Gagnon
Background Dementia is one of the main public health priorities for current and future societies worldwide. Over the past years, eHealth sol… (see more)utions have added numerous promising solutions to enhance the health and wellness of people living with dementia-related cognitive problems and their primary caregivers. Previous studies have shown that an environmental scan identifies the knowledge-to-action gap meaningfully. This paper presents the protocol of an environmental scan to monitor the currently available eHealth solutions targeting dementia and other neurocognitive disorders against selected attributes. Objective This study aims to identify the characteristics of currently available eHealth solutions recommended for older adults with cognitive problems and their informal caregivers. To inform the recommendations regarding eHealth solutions for these people, it is important to obtain a comprehensive view of currently available technologies and document their outcomes and conditions of success. Methods We will perform an environmental scan of available eHealth solutions for older adults with cognitive impairment or dementia and their informal caregivers. Potential solutions will be initially identified from a previous systematic review. We will also conduct targeted searches for gray literature on Google and specialized websites covering the regions of Canada and Europe. Technological tools will be scanned based on a preformatted extraction grid. The relevance and efficiency based on the selected attributes will be assessed. Results We will prioritize relevant solutions based on the needs and preferences identified from a qualitative study among older adults with cognitive impairment or dementia and their informal caregivers. Conclusions This environmental scan will identify eHealth solutions that are currently available and scientifically appraised for older adults with cognitive impairment or dementia and their informal caregivers. This knowledge will inform the development of a decision support tool to assist older adults and their informal caregivers in their search for adequate eHealth solutions according to their needs and preferences based on trustable information. International Registered Report Identifier (IRRID) DERR1-10.2196/41015
2022-11-04
JMIR Research Protocols (published)
doi.org
Representational ethical model calibration
Robert Carruthers
Isabel Straw
James K. Ruffle
Daniel Herron
Amy Nelson
Danilo Bzdok
Delmiro Fernandez-Reyes
Geraint Rees
Parashkev Nachev
2022-11-04
NPJ Digital Medicine (published)
doi.org
arxiv.org
Spectral Regularization: an Inductive Bias for Sequence Modeling
Kaiwen Hou
Guillaume Rabusseau
Hou Rabusseau
2022-11-04
ArXiv (preprint)
doi.org
arxiv.org
Adult neurogenesis acts as a neural regularizer
Lina M. Tran
Adam Santoro
Lulu Liu
Sheena A. Josselyn
Blake Richards
Paul W. Frankland
2022-11-02
Proceedings of the National Academy of Sciences of the United States of America (published)
doi.org
Global SARS-CoV-2 seroprevalence from January 2020 to April 2022: A systematic review and meta-analysis of standardized population-based studies
Isabel Bergeri
Mairead Whelan
Harriet Ware
Lorenzo Subissi
Anthony Nardone
Hannah C. Lewis
Zihan Li
Xiaomeng Ma
Marta Valenciano
Brianna Cheng
Lubna Al Ariqi
Arash Rashidian
Joseph Okeibunor
Tasnim Azim
Pushpa Wijesinghe
Linh-Vi Le
Aisling Vaughan
Richard Pebody
Andrea Vicari
Tingting Yan … (see 9 more)
Mercedes Yanes-Lane
Christian Cao
David A. Clifton
Matthew P. Cheng
Jesse Papenburg
David Buckeridge
Niklas Bobrovitz
Rahul K. Arora
Maria D Van Kerkhove
2022-11-01
PLoS Medicine (published)
doi.org
Successive-Cancellation Decoding of Reed-Muller Codes With Fast Hadamard Transform
Nghia Doan
Seyyed Ali Hashemi
Warren Gross
A novel permuted fast successive-cancellation list decoding algorithm with fast Hadamard transform (FHT-FSCL) is presented. The proposed dec… (see more)oder initializes
2022-11-01
IEEE Transactions on Vehicular Technology (published)
doi.org
arxiv.org
Rethinking Generalization: The Impact of Annotation Style on Medical Image Segmentation
Brennan Nichyporuk
Jillian L. Cardinell
Justin Szeto
Raghav Mehta
Jean-Pierre R. Falet
Douglas Arnold
Sotirios A. Tsaftaris
Tal Arbel
Generalization is an important attribute of machine learning models, particularly for those that are to be deployed in a medical context, wh… (see more)ere unreliable predictions can have real world consequences. While the failure of models to generalize across datasets is typically attributed to a mismatch in the data distributions, performance gaps are often a consequence of biases in the "ground-truth" label annotations. This is particularly important in the context of medical image segmentation of pathological structures (e.g. lesions), where the annotation process is much more subjective, and affected by a number underlying factors, including the annotation protocol, rater education/experience, and clinical aims, among others. In this paper, we show that modeling annotation biases, rather than ignoring them, poses a promising way of accounting for differences in annotation style across datasets. To this end, we propose a generalized conditioning framework to (1) learn and account for different annotation styles across multiple datasets using a single model, (2) identify similar annotation styles across different datasets in order to permit their effective aggregation, and (3) fine-tune a fully trained model to a new annotation style with just a few samples. Next, we present an image-conditioning approach to model annotation styles that correlate with specific image features, potentially enabling detection biases to be more easily identified.
2022-10-31
ArXiv (preprint)
doi.org
arxiv.org