Publications (19)
Asynchronous Gossip Algorithms for Rank-Based Statistical Methods
Anna Van Elst, Igor Colin, Stephan Clémençon
As decentralized AI and edge intelligence become increasingly prevalent, ensuring robustness and trustworthiness in such distributed settings has become a critical issue-especially…
Parallel Contextual Bandits in Wireless Handover Optimization
Igor Colin, Albert Thomas, Moez Draief
As cellular networks become denser, a scalable and dynamic tuning of wireless base station parameters can only be achieved through automated optimization. Although the contextual b…
Adaptive Sample Sharing for Multi Agent Linear Bandits
Hamza Cherkaoui, Merwan Barlier, Igor Colin
The multi-agent linear bandit setting is a well-known setting for which designing efficient collaboration between agents remains challenging. This paper studies the impact of data…
Refined bounds for randomized experimental design
Geovani Rizk, Igor Colin, Albert Thomas +1
Experimental design is an approach for selecting samples among a given set so as to obtain the best estimator for a given criterion. In the context of linear regression, several op…
Scaling-up Empirical Risk Minimization: Optimization of Incomplete U-statistics
Stéphan Clémençon, Aurélien Bellet, Igor Colin
In a wide range of statistical learning problems such as ranking, clustering or metric learning among others, the risk is accurately estimated by -statistics of degree …
On Gossip Algorithms for Machine Learning with Pairwise Objectives
Igor Colin, Aurélien Bellet, Stephan Clémençon +1
In the IoT era, information is more and more frequently picked up by connected smart sensors with increasing, though limited, storage, communication and computation abilities. Whet…
Price of Safety in Linear Best Arm Identification
Xuedong Shang, Igor Colin, Merwan Barlier +1
We introduce the safe best-arm identification framework with linear feedback, where the agent is subject to some stage-wise safety constraint that linearly depends on an unknown pa…
Decentralized Topic Modelling with Latent Dirichlet Allocation
Igor Colin, Christophe Dupuy
Privacy preserving networks can be modelled as decentralized networks (e.g., sensors, connected objects, smartphones), where communication between nodes of the network is not contr…
Theoretical Limits of Pipeline Parallel Optimization and Application to Distributed Deep Learning
Igor Colin, Ludovic Dos Santos, Kevin Scaman
We investigate the theoretical limits of pipeline parallel learning of deep learning architectures, a distributed setup in which the computation is distributed per layer instead of…
Gossip Dual Averaging for Decentralized Optimization of Pairwise Functions
Igor Colin, Aurélien Bellet, Joseph Salmon +1
In decentralized networks (of sensors, connected objects, etc.), there is an important need for efficient algorithms to optimize a global cost function, for instance to learn a glo…
Robust Distributed Estimation: Extending Gossip Algorithms to Ranking and Trimmed Means
Anna Van Elst, Igor Colin, Stephan Clémençon
This paper addresses the problem of robust estimation in gossip algorithms over arbitrary communication graphs. Gossip algorithms are fully decentralized, relying only on local nei…
Differentially Private Policy Gradient
Alexandre Rio, Merwan Barlier, Igor Colin
Motivated by the increasing deployment of reinforcement learning in the real world, involving a large consumption of personal data, we introduce a differentially private (DP) polic…
An Approximate Shapley-Folkman Theorem
Thomas Kerdreux, Igor Colin, Alexandre d'Aspremont
The Shapley-Folkman theorem shows that Minkowski averages of uniformly bounded sets tend to be convex when the number of terms in the sum becomes much larger than the ambient dimen…
Differentially Private Deep Model-Based Reinforcement Learning
Alexandre Rio, Merwan Barlier, Igor Colin +1
We address private deep offline reinforcement learning (RL), where the goal is to train a policy on standard control tasks that is differentially private (DP) with respect to indiv…
An -No-Regret Algorithm For Graphical Bilinear Bandits
Geovani Rizk, Igor Colin, Albert Thomas +2
We propose the first regret-based approach to the Graphical Bilinear Bandits problem, where agents in a graph play a stochastic bilinear bandit game with each of their neighbor…
Decentralized Ranking Aggregation via Gossip: Convergence and Robustness
Kerrian Le Caillec, Anna Van Elst, Igor Colin +1
The concept of ranking aggregation plays a central role in preference analysis, and numerous algorithms for calculating median rankings, often originating in social choice theory,…
Fast and Efficient Gossip Algorithms for Robust and Non-smooth Decentralized Learning
Anna van Elst, Igor Colin, Stephan Clémençon
Decentralized learning on resource-constrained edge devices demands algorithms that are communication-efficient, robust to data corruption, and lightweight in memory. State-of-the-…
Extending Gossip Algorithms to Distributed Estimation of U-Statistics
Igor Colin, Aurélien Bellet, Joseph Salmon +1
Efficient and robust algorithms for decentralized estimation in networks are essential to many distributed systems. Whereas distributed estimation of sample mean statistics has bee…
Best Arm Identification in Graphical Bilinear Bandits
Geovani Rizk, Albert Thomas, Igor Colin +2
We introduce a new graphical bilinear bandit problem where a learner (or a \emph{central entity}) allocates arms to the nodes of a graph and observes for each edge a noisy bilinear…