[Home]
[
News]
[ Events]
[ People]
[ Research]
[ Education]
[Visitor Info]
[UCSD Only]
[Admin]
Adaptive Cubic Regularization Methods for Nonconvex Unconstrained Optimization
Ziyan Zhu
UCSD
Abstract:
Adaptive cubic regularization methods have several favorable properties for nonconvex optimization. In particular, under mild assumptions, they are globally convergent to a second-order stationary point. In this talk, I will introduce an adaptive cubic regularization method for unconstrained optimization. Methods analogous to those used to solve the trust-region subproblem will be discussed for solving the local cubic model. Some numerical results will be presented that compare a cubic regularized Newton's method, a standard trust-region method and a trust-search method.
Tuesday, November 12, 2019
11:00AM AP&M 2402
News]
[ Events]
[ People]
[ Research]
[ Education]
[Visitor Info]
[UCSD Only]
[Admin]
Directors:
Randolph E. Bank
Philip E. Gill
Michael Holst
Administrative Contact:
[ Click to Send Email ]
Randolph E. Bank
Philip E. Gill
Michael Holst
Administrative Contact:
[ Click to Send Email ]
Ziyan Zhu
UCSD
Abstract:
Adaptive cubic regularization methods have several favorable properties for nonconvex optimization. In particular, under mild assumptions, they are globally convergent to a second-order stationary point. In this talk, I will introduce an adaptive cubic regularization method for unconstrained optimization. Methods analogous to those used to solve the trust-region subproblem will be discussed for solving the local cubic model. Some numerical results will be presented that compare a cubic regularized Newton's method, a standard trust-region method and a trust-search method.
Tuesday, November 12, 2019
11:00AM AP&M 2402