OCLGMLFeb 15, 2018

A Progressive Batching L-BFGS Method for Machine Learning

arXiv:1802.05374v2174 citations
Originality Incremental advance
AI Analysis

This incremental method addresses the problem of slow convergence and poor generalization in L-BFGS for machine learning practitioners.

The authors tackled the challenge of making L-BFGS effective for large-scale machine learning by introducing a progressive batching approach that increases sample size during optimization, resulting in improved performance on logistic regression and deep neural networks.

The standard L-BFGS method relies on gradient approximations that are not dominated by noise, so that search directions are descent directions, the line search is reliable, and quasi-Newton updating yields useful quadratic models of the objective function. All of this appears to call for a full batch approach, but since small batch sizes give rise to faster algorithms with better generalization properties, L-BFGS is currently not considered an algorithm of choice for large-scale machine learning applications. One need not, however, choose between the two extremes represented by the full batch or highly stochastic regimes, and may instead follow a progressive batching approach in which the sample size increases during the course of the optimization. In this paper, we present a new version of the L-BFGS algorithm that combines three basic components - progressive batching, a stochastic line search, and stable quasi-Newton updating - and that performs well on training logistic regression and deep neural networks. We provide supporting convergence theory for the method.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes