Geometry Based Acceleration for Non-smooth Optimisation

Add to your list(s) Download to your calendar using vCal

• Dr. Jingwei Liang, Cambridge University
• Monday 28 October 2019, 14:00-15:00
• Room 909B, EEE Building.

If you have a question about this talk, please contact Wei Dai.

Abstract: The acceleration of first-order methods for non-smooth optimization is an active research area, due to the popularity of first-order methods. In this talk, I will present a geometry based adaptive acceleration framework for speeding up general first-order methods when applied to non-smooth optimisation. Such a framework can render algorithms which are able to automatically adjust themselves to the underlying geometry of the optimisation problems and the structure of the first-order methods. Connection with vector extrapolation will be discussed. State-of-the-art performances are obtained on various non-smooth optimisation problems and first-order methods, particularly the non-descent type methods.

Short bio: Jingwei Liang is now a Leverhulme Early Career Fellowship at DAMTP University of Cambridge where he also spent two years at Postdoc Research Associate. He received his Ph.D.candidate in applied mathematics in 2016 from ENSI -CAEN and University of Normandy France. Before that, he received his master degree in applied mathematics in 2013 from Shanghai Jiao Tong University, China, and bachelor degree in Electronic and InformationEngineering from Nanjing University of Posts and Telecommunications. His current research interests mainly focus on optimization.

This talk is part of the COMMSP Seminar series.

This talk is included in these lists:

• COMMSP & CP list
• COMMSP Seminar
• Room 909B, EEE Building
• SAP Talks

Note that ex-directory lists are not shown.