A Grothendieck-type inequality for local maxima

03/13/2016
by   Andrea Montanari, et al.
0

A large number of problems in optimization, machine learning, signal processing can be effectively addressed by suitable semidefinite programming (SDP) relaxations. Unfortunately, generic SDP solvers hardly scale beyond instances with a few hundreds variables (in the underlying combinatorial problem). On the other hand, it has been observed empirically that an effective strategy amounts to introducing a (non-convex) rank constraint, and solving the resulting smooth optimization problem by ascent methods. This non-convex problem has --generically-- a large number of local maxima, and the reason for this success is therefore unclear. This paper provides rigorous support for this approach. For the problem of maximizing a linear functional over the elliptope, we prove that all local maxima are within a small gap from the SDP optimum. In several problems of interest, arbitrarily small relative error can be achieved by taking the rank constraint k to be of order one, independently of the problem size.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
03/25/2017

Solving SDPs for synchronization and MaxCut problems via the Grothendieck inequality

A number of statistical estimation problems can be addressed by semidefi...
research
02/17/2017

Solving Equations of Random Convex Functions via Anchored Regression

We consider the question of estimating a solution to a system of equatio...
research
02/01/2016

Solving rank-constrained semidefinite programs in exact arithmetic

We consider the problem of minimizing a linear function over an affine s...
research
06/16/2021

Momentum-inspired Low-Rank Coordinate Descent for Diagonally Constrained SDPs

We present a novel, practical, and provable approach for solving diagona...
research
07/12/2018

Convergence Rate of Block-Coordinate Maximization Burer-Monteiro Method for Solving Large SDPs

Semidefinite programming (SDP) with equality constraints arise in many o...
research
10/16/2020

Strengthened SDP Verification of Neural Network Robustness via Non-Convex Cuts

There have been major advances on the design of neural networks, but sti...
research
10/15/2022

Bearing-based Relative Localization for Robotic Swarm with Partially Mutual Observations

Mutual localization provides a consensus of reference frame as an essent...

Please sign up or login with your details

Forgot password? Click here to reset