How to solve a classification problem using a cooperative tiling Multi-Agent System?

Comment résoudre un problème de classification avec un système multi-agent coopératif ?

Thibault Fourez (1, 2) , Nicolas Verstaevel (2) , Frédéric Migeon (2) , Frédéric Schettini (3) , Frederic Amblard (2)

1 Citec Ingénieurs Conseils SAS
2 IRIT-SMAC - Systèmes Multi-Agents Coopératifs
3 Citec Ingénieurs Conseils SAS

Adaptive Multi-Agent Systems (AMAS) transform dynamic problems into problems of local cooperation between agents. We present smapy, an ensemble based AMAS implementation for mobility prediction, whose agents are provided with machine learning models in addition to their cooperation rules. With a detailed methodology, we propose a framework to transform a classification problem into a cooperative tiling of the input variable space. We show that it is possible to use linear classifiers for online non-linear classification on three benchmark toy problems chosen for their different levels of linear separability, if they are integrated in a cooperative Multi-Agent structure. The results obtained show a significant improvement of the performance of linear classifiers in non-linear contexts in terms of classification accuracy and decision boundaries, thanks to the cooperative approach.

Keywords: Adaptative Multi-Agent System; Ensemble learning; Nonlinear classification

Domains: Artificial Intelligence [cs.AI]; Multiagent Systems [cs.MA]; Machine Learning [stat.ML]

Thibault Fourez, Nicolas Verstaevel, Frédéric Migeon, Frédéric Schettini, Frederic Amblard. How to solve a classification problem using a cooperative tiling Multi-Agent System?. 20th International Conference on Practical Applications of Agents and Multi-Agent Systems (PAAMS 2022), Jul 2022, L'Aquila, Italy. à paraître. ⟨hal-03776374⟩