University of Cambridge > Talks.cam > CUED Control Group Seminars > Stochastic Algorithms for Nonconvex Optimization

  Browse A-Z

Log in

University Account External (via Google)

🔐 Log In

Information on

Subscribing to talksdetails Finding a talkdetails Adding a talkdetails Disseminating talksdetails Help and Documentationdetails

ℹ️ Information

Stochastic Algorithms for Nonconvex Optimization

Download to your calendar using vCal

If you have a question about this talk, please contact Fulvio Forni .

Nowadays it is quite common to solve optimization problems in 10^9 or more variables. At these levels, it is not practical to use the “true” gradient of the objective function. Instead, a variety of methods are based on using approximate gradients, which are also random; in other words, they are stochastic gradients. Often, only some components of the argument are updated at each iteration, to reduce storage calls. As a result, nowadays optimization algorithms produce stochastic processes, as opposed to sequences of vectors in some Euclidean space. Another issue is that the objective function is not convex.

The seminar will be held in LR3A , Department of Engineering, and online (zoom): https://newnham.zoom.us/j/92544958528?pwd=YS9PcGRnbXBOcStBdStNb3E0SHN1UT09

This talk is part of the CUED Control Group Seminars series.

This talk is included in these lists:

Note that ex-directory lists are not shown.

Other lists

National Centre for Statistical Ecology (NCSE) Seminars dress design Wolfson Lecture Theatre Yusuf Hamied Department of Chemistry

Other talks

Title to be confirmed Cooperation for Nature Haematology and Palliative Care Emergency Medicine and Metabolic Medicine Title to be confirmed What’s my reproduction number? Life history strategies: an interdisciplinary perspective
 

© 2006-2025 Talks.cam, University of Cambridge. Contact Us | Help and Documentation | Privacy and Publicity