Decentralized Collaborative Learning of Personalized Models over Networks

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