.png){kind=icon}
.png){kind=icon}
Return
SEARCH DIRECTION CORRECTION WITH NORMALIZED GRADIENT MAKES FIRST-ORDER METHODS FASTER
.png){kind=icon}
PRE
AI
DOI：10.1137/20M1335480.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
The so-called fast inertial relaxation engine is a first-order method for unconstrained smooth optimization problems. It updates the search direction by a linear combination of the past search direction, the current gradient, and the normalized gradient direction. We explore more general combination rules and call this generalized technique the search direction correction (SDC). SDC is extended to composite and stochastic optimization problems as well. Deriving from a second order ODE, we propose a fast inertial search direction correction (FISC) algorithm as an example of methods with SDC. We prove the O(k(-2)) convergence rate of FISC for convex optimization problems. Numerical results on sparse optimization, logistic regression, as well as deep learning demonstrate that our proposed methods are quite competitive to other state-of-the-art first-order algorithms.
Keywords:
first-order methods
search direction correction
Lyapunov function
composite optimization
stochastic optimization
Journal
IF:
2.6
Papers:
5.1K
Citations:
1.8W