Pascal Germain

Membre académique associé

Professeur adjoint, Université Laval, Département d'informatique et de génie logiciel

Sujets de recherche

Apprentissage de représentations

Théorie de l'apprentissage automatique

Site web

Google Scholar

Biographie

Pascal Germain est professeur adjoint au Département d’informatique et de génie logiciel de l’Université Laval et chercheur en apprentissage automatique. Il a obtenu un doctorat en informatique de l'Université Laval en 2015, sous la direction de François Laviolette et de Mario Marchand. Il a ensuite poursuivi ses travaux de recherche en France pendant quatre ans au sein de l’Institut national de recherche en sciences et technologies du numérique (Inria), d'abord comme postdoctorant dans l'équipe du projet SIERRA de Francis Bach, puis comme chargé de recherche et membre de l'équipe du projet MODAL. Il a aussi été membre affilié et enseignant au Département de mathématiques de l'Université de Lille. De retour à son alma mater en tant que professeur adjoint depuis 2019, il y est membre du Centre de recherche en données massives (CRDM) et de l'Institut intelligence et données (IID). Il est également membre académique associé de Mila – Institut québécois d’intelligence artificielle. Ses domaines de recherche comprennent la théorie statistique de l’apprentissage automatique, l'apprentissage par transfert et l'apprentissage de prédicteurs interprétables.

Étudiants actuels

Mathieu Bazinet

Doctorat - Université Laval

Github

Benjamin Leblanc

Doctorat - Université Laval

Site web

Google Scholar

Publications

On Selecting Robust Approaches for Learning Predictive Biomarkers in Metabolomics Data Sets.

Thibaud Godon

Pier-Luc Plante

Jacques Corbeil

Pascal Germain

Alexandre Drouin

Metabolomics, the study of small molecules within biological systems, offers insights into metabolic processes and, consequently, holds grea… (voir plus)t promise for advancing health outcomes. Biomarker discovery in metabolomics represents a significant challenge, notably due to the high dimensionality of the data. Recent work has addressed this problem by analyzing the most important variables in machine learning models. Unfortunately, this approach relies on prior hypotheses about the structure of the data and may overlook simple patterns. To assess the true usefulness of machine learning methods, we evaluate them on a collection of 835 metabolomics data sets. This effort provides valuable insights for metabolomics researchers regarding where and when to use machine learning. It also establishes a benchmark for the evaluation of future methods. Nonetheless, the results emphasize the high diversity of data sets in metabolomics and the complexity of finding biologically relevant biomarkers. As a result, we propose a novel approach applicable across all data sets, offering guidance for future analyses. This method involves directly comparing univariate and multivariate models. We demonstrate through selected examples how this approach can guide data analysis across diverse data set structures, representative of the observed variability. Code and data are available for research purposes.

2025-06-12

Analytical Chemistry (publié)

doi.org

Sample Compression for Continual Learning

Jacob Comeau

Mathieu Bazinet

Pascal Germain

Cem Subakan

2025-03-13

ArXiv (prépublication)

doi.org

arxiv.org

Sample Compression for Continual Learning

Jacob Comeau

Mathieu Bazinet

Pascal Germain

Cem Subakan

2025-03-13

ArXiv (prépublication)

arxiv.org

Sample Compression for Self Certified Continual Learning

Jacob Comeau

Mathieu Bazinet

Pascal Germain

Cem Subakan

Continual learning algorithms aim to learn from a sequence of tasks, making the training distribution non-stationary. The majority of existi… (voir plus)ng continual learning approaches in the literature rely on heuristics and do not provide learning guarantees. In this paper, we present a new method called Continual Pick-to-Learn (CoP2L), which is able to retain the most representative samples for each task in an efficient way. CoP2L combines the Pick-to-Learn algorithm (rooted in the sample compression theory) and the experience replay continual learning scheme. This allows us to provide non-vacuous upper bounds on the generalization loss of the learned predictors, numerically computable after each task. We empirically evaluate our approach on several standard continual learning benchmarks across Class-Incremental, Task-Incremental, and Domain-Incremental settings. Our results show that CoP2L is highly competitive across all setups, often outperforming existing baselines, and significantly mitigating catastrophic forgetting compared to vanilla experience replay in the Class-Incremental setting. It is possible to leverage the bounds provided by CoP2L in practical scenarios to certify the predictor reliability on previously learned tasks, in order to improve the trustworthiness of the continual learning algorithm.

2025-03-13

ArXiv (prépublication)

arxiv.org

Sample compression unleashed : New generalization bounds for real valued losses

Mathieu Bazinet

Valentina Zantedeschi

Pascal Germain

2025-01-22

aistats.org/AISTATS/2025/Conference (poster)

openreview.net

Sample Compression Hypernetworks: From Generalization Bounds to Meta-Learning

Benjamin Leblanc

Mathieu Bazinet

Nathaniel D'Amours

Alexandre Drouin

Pascal Germain

Reconstruction functions are pivotal in sample compression theory, a framework for deriving tight generalization bounds. From a small sample… (voir plus) of the training set (the compression set) and an optional stream of information (the message), they recover a predictor previously learned from the whole training set. While usually fixed, we propose to learn reconstruction functions. To facilitate the optimization and increase the expressiveness of the message, we derive a new sample compression generalization bound for real-valued messages. From this theoretical analysis, we then present a new hypernetwork architecture that outputs predictors with tight generalization guarantees when trained using an original meta-learning framework. The results of promising preliminary experiments are then reported.

2024-10-09

NeurIPS.cc/2024/Workshop/Compression (publié)

openreview.net