Title: Optimization in Data Analysis: Some Recent Developments
| Speaker: | Stephen J. Wright |
| Affiliation: |
Computer Sciences Department and Wisconsin Institute for Discovery University of Wisconsin-Madison |
| Room: | MC 5501 |
Abstract:
Optimization is vital to the modern revolution in data science, and techniques from optimization have become essential in formulating and solving a wide variety of data analysis problems. In turn, data science has caused a ferment of new research activity in optimization by posing challenging new problems and new contexts. We start this talk with an overview of the many problem classes in data science in which optimization provides the key solution methodology. We then focus on several areas of growing recent interest, including the interplay between optimization and data analysis in such areas as nonconvex optimization, robust optimization, adversarial machine learning, neural networks, and matrix optimization.