Publications de Youssef Allouah

The Utility and Complexity of In- and Out-of-Distribution Machine Unlearning

Machine unlearning, the process of selectively removing data from trained models, is increasingly crucial for addressing privacy concerns and knowledge gaps post-deployment. Despite this importance, existing approaches are often heuristic and lack formal g ...

ICLR2025

Robust Sparse Voting

Rachid Guerraoui, Youssef Allouah, Oscar Jean Olivier Villemaud, Le Nguyen Hoang

Many applications, such as content moderation and recommendation, require reviewing and scoring a large number of alternatives. Doing so robustly is however very challenging. Indeed, voters' inputs are inevitably sparse: most alternatives are only scored b ...

2024

Revisiting Ensembling in One-Shot Federated Learning

Rachid Guerraoui, Anne-Marie Kermarrec, Rafael Pereira Pires, Youssef Allouah, Akash Balasaheb Dhasade, Rishi Sharma, Nirupam Gupta, Rafaël Benjamin Pinot

Federated learning (FL) is an appealing approach to training machine learning models without sharing raw data. However, standard FL algorithms are iterative and thus induce a significant communication cost. One-shot federated learning (OFL) trades the iter ...

Curran Associates, Inc.2024

Fine-Tuning Personalization in Federated Learning to Mitigate Adversarial Clients

Rachid Guerraoui, Youssef Allouah, Nirupam Gupta, Rafaël Benjamin Pinot

Federated learning (FL) is an appealing paradigm that allows a group of machines (a.k.a. clients) to learn collectively while keeping their data local. However, due to the heterogeneity between the clients' data distributions, the model obtained through th ...

Neural information processing systems foundation2024

The Privacy Power of Correlated Noise in Decentralized Learning

Rachid Guerraoui, Martin Jaggi, Youssef Allouah, Anastasiia Koloskova, Aymane El Firdoussi

Decentralized learning is appealing as it enables the scalable usage of large amounts of distributed data and resources (without resorting to any central entity), while promoting privacy since every user minimizes the direct exposure of their data. Yet, wi ...

PMLR2024

On the Privacy-Robustness-Utility Trilemma in Distributed Learning

Rachid Guerraoui, John Stephan, Youssef Allouah, Nirupam Gupta, Rafaël Benjamin Pinot

The ubiquity of distributed machine learning (ML) in sensitive public domain applications calls for algorithms that protect data privacy, while being robust to faults and adversarial behaviors. Although privacy and robustness have been extensively studied ...

2023

Robust Distributed Learning: Tight Error Bounds and Breakdown Point under Data Heterogeneity

Rachid Guerraoui, Youssef Allouah, Geovani Rizk, Nirupam Gupta, Rafaël Benjamin Pinot

The theory underlying robust distributed learning algorithms, designed to resist adversarial machines, matches empirical observations when data is homogeneous. Under data heterogeneity however, which is the norm in practical scenarios, established lower bo ...

2023

Fixing by Mixing: A Recipe for Optimal Byzantine ML under Heterogeneity

Rachid Guerraoui, John Stephan, Sadegh Farhadkhani, Youssef Allouah, Nirupam Gupta, Rafaël Benjamin Pinot

Byzantine machine learning (ML) aims to ensure the resilience of distributed learning algorithms to misbehaving (or Byzantine) machines. Although this problem received significant attention, prior works often assume the data held by the machines to be homo ...

PMLR2023

Further results on latent discourse models and word embeddings

Youssef Allouah

We discuss some properties of generative models for word embeddings. Namely, (Arora et al., 2016) proposed a latent discourse model implying the concentration of the partition function of the word vectors. This concentration phenomenon led to an asymptotic ...

Microtome Publishing2021