Bootstrap Model Aggregation for Distributed Statistical Learning

TL;DR

This paper introduces variance reduction techniques for bootstrap-based model aggregation in distributed learning, improving statistical efficiency and robustness when combining models from different sources.

Contribution

It proposes two novel variance reduction methods, including a weighted M-estimator, to enhance bootstrap model aggregation in distributed statistical learning.

Findings

The methods reduce bootstrap noise effectively.

The approaches are both theoretically sound and empirically validated.

Abstract

In distributed, or privacy-preserving learning, we are often given a set of probabilistic models estimated from different local repositories, and asked to combine them into a single model that gives efficient statistical estimation. A simple method is to linearly average the parameters of the local models, which, however, tends to be degenerate or not applicable on non-convex models, or models with different parameter dimensions. One more practical strategy is to generate bootstrap samples from the local models, and then learn a joint model based on the combined bootstrap set. Unfortunately, the bootstrap procedure introduces additional noise and can significantly deteriorate the performance. In this work, we propose two variance reduction methods to correct the bootstrap noise, including a weighted M-estimator that is both statistically efficient and practically powerful. Both theoretical and empirical analysis is provided to demonstrate our methods.