Decentralized Collaborative Learning of Personalized Models over Networks

Paul Vanhaesebrouck (1) , Aurélien Bellet (1) , Marc Tommasi (1, 2)

1 MAGNET - Machine Learning in Information Networks
2 Université de Lille

We consider a set of learning agents in a col-laborative peer-to-peer network, where each agent learns a personalized model according to its own learning objective. The question addressed in this paper is: how can agents improve upon their locally trained model by communicating with other agents that have similar objectives? We introduce and analyze two asynchronous gossip algorithms running in a fully decentralized manner. Our first approach , inspired from label propagation, aims to smooth pre-trained local models over the network while accounting for the confidence that each agent has in its initial model. In our second approach, agents jointly learn and propagate their model by making iterative updates based on both their local dataset and the behavior of their neighbors. To optimize this challenging objective, our decentralized algorithm is based on ADMM.

Domaines: Machine Learning [stat.ML]; Intelligence artificielle [cs.AI]; Calcul parallèle, distribué et partagé [cs.DC]; Systèmes et contrôle [cs.SY]

Paul Vanhaesebrouck, Aurélien Bellet, Marc Tommasi. Decentralized Collaborative Learning of Personalized Models over Networks. International Conference on Artificial Intelligence and Statistics (AISTATS), Apr 2017, Fort Lauderdale, Florida., United States. ⟨hal-01533182⟩