Gossiped and Quantized Online Multi-Kernel Learning

TL;DR

This paper extends online multi-kernel learning to non-fully connected networks using a gossip algorithm, demonstrating sub-linear regret and effective communication management in wireless sensor networks.

Contribution

It introduces a gossip-based algorithm for distributed online multi-kernel learning on non-fully connected graphs, with proven sub-linear regret.

Findings

Gossip algorithm achieves sub-linear regret in non-fully connected networks.

Quantized communication reduces load without sacrificing learning performance.

Experimental results validate theoretical claims on real datasets.

Abstract

In instances of online kernel learning where little prior information is available and centralized learning is unfeasible, past research has shown that distributed and online multi-kernel learning provides sub-linear regret as long as every pair of nodes in the network can communicate (i.e., the communications network is a complete graph). In addition, to manage the communication load, which is often a performance bottleneck, communications between nodes can be quantized. This letter expands on these results to non-fully connected graphs, which is often the case in wireless sensor networks. To address this challenge, we propose a gossip algorithm and provide a proof that it achieves sub-linear regret. Experiments with real datasets confirm our findings.