Pith. sign in

REVIEW

Linear Speedup in Saddle-Point Escape for Decentralized Non-Convex Optimization

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1910.13852 v1 pith:YJPJ2WIR submitted 2019-10-30 cs.MA cs.LGmath.OCstat.ML

classification cs.MAcs.LGmath.OCstat.ML
keywords linearspeedupcombinationnon-convexoptimizationagentsbeendecentralized
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Under appropriate cooperation protocols and parameter choices, fully decentralized solutions for stochastic optimization have been shown to match the performance of centralized solutions and result in linear speedup (in the number of agents) relative to non-cooperative approaches in the strongly-convex setting. More recently, these results have been extended to the pursuit of first-order stationary points in non-convex environments. In this work, we examine in detail the dependence of second-order convergence guarantees on the spectral properties of the combination policy for non-convex multi agent optimization. We establish linear speedup in saddle-point escape time in the number of agents for symmetric combination policies and study the potential for further improvement by employing asymmetric combination weights. The results imply that a linear speedup can be expected in the pursuit of second-order stationary points, which exclude local maxima as well as strict saddle-points and correspond to local or even global minima in many important learning settings.

Discussion (0). Sign in to comment.

Pith tools