Third-order Smoothness Helps: Even Faster Stochastic Optimization Algorithms for Finding Local Minima

by   Yaodong Yu, et al.

We propose stochastic optimization algorithms that can find local minima faster than existing algorithms for nonconvex optimization problems, by exploiting the third-order smoothness to escape non-degenerate saddle points more efficiently. More specifically, the proposed algorithm only needs Õ(ϵ^-10/3) stochastic gradient evaluations to converge to an approximate local minimum x, which satisfies ∇ f(x)_2≤ϵ and λ_(∇^2 f(x))≥ -√(ϵ) in the general stochastic optimization setting, where Õ(·) hides logarithm polynomial terms and constants. This improves upon the Õ(ϵ^-7/2) gradient complexity achieved by the state-of-the-art stochastic local minima finding algorithms by a factor of Õ(ϵ^-1/6). For nonconvex finite-sum optimization, our algorithm also outperforms the best known algorithms in a certain regime.


page 1

page 2

page 3

page 4


Finding Local Minima via Stochastic Nested Variance Reduction

We propose two algorithms that can find local minima faster than the sta...

Faster Perturbed Stochastic Gradient Methods for Finding Local Minima

Escaping from saddle points and finding local minima is a central proble...

Saving Gradient and Negative Curvature Computations: Finding Local Minima More Efficiently

We propose a family of nonconvex optimization algorithms that are able t...

Neon2: Finding Local Minima via First-Order Oracles

We propose a reduction for non-convex optimization that can (1) turn a s...

A Note On The Popularity of Stochastic Optimization Algorithms in Different Fields: A Quantitative Analysis from 2007 to 2017

Stochastic optimization algorithms are often used to solve complex large...

Learning to Optimize Neural Nets

Learning to Optimize is a recently proposed framework for learning optim...

Follow the Signs for Robust Stochastic Optimization

Stochastic noise on gradients is now a common feature in machine learnin...