Log In Sign Up

Escaping Saddle Points with Stochastically Controlled Stochastic Gradient Methods

by   Guannan Liang, et al.

Stochastically controlled stochastic gradient (SCSG) methods have been proved to converge efficiently to first-order stationary points which, however, can be saddle points in nonconvex optimization. It has been observed that a stochastic gradient descent (SGD) step introduces anistropic noise around saddle points for deep learning and non-convex half space learning problems, which indicates that SGD satisfies the correlated negative curvature (CNC) condition for these problems. Therefore, we propose to use a separate SGD step to help the SCSG method escape from strict saddle points, resulting in the CNC-SCSG method. The SGD step plays a role similar to noise injection but is more stable. We prove that the resultant algorithm converges to a second-order stationary point with a convergence rate of Õ( ϵ^-2 log( 1/ϵ)) where ϵ is the pre-specified error tolerance. This convergence rate is independent of the problem dimension, and is faster than that of CNC-SGD. A more general framework is further designed to incorporate the proposed CNC-SCSG into any first-order method for the method to escape saddle points. Simulation studies illustrate that the proposed algorithm can escape saddle points in much fewer epochs than the gradient descent methods perturbed by either noise injection or a SGD step.


page 1

page 2

page 3

page 4


Escaping Saddles with Stochastic Gradients

We analyze the variance of stochastic gradients along negative curvature...

Escaping Saddle Points with Adaptive Gradient Methods

Adaptive methods such as Adam and RMSProp are widely used in deep learni...

Stopping Criteria for, and Strong Convergence of, Stochastic Gradient Descent on Bottou-Curtis-Nocedal Functions

While Stochastic Gradient Descent (SGD) is a rather efficient algorithm ...

Stochastic Subgradient Descent Escapes Active Strict Saddles

In non-smooth stochastic optimization, we establish the non-convergence ...

Bolstering Stochastic Gradient Descent with Model Building

Stochastic gradient descent method and its variants constitute the core ...

NeCPD: An Online Tensor Decomposition with Optimal Stochastic Gradient Descent

Multi-way data analysis has become an essential tool for capturing under...