- Seminar Calendar
- Seminar Archive
- 2024-2025 Semester 2
- 2024-2025 Semester 1
- 2023-2024 Semester 2
- 2023-2024 Semester 1
- 2022-2023 Semester 2
- 2022-2023 Semester 1
- 2021-2022 Semester 2
- 2021-2022 Semester 1
- 2020-2021 Semester 2
- 2020-2021 Semester 1
- 2019-2020 Semester 2
- 2019-2020 Semester 1
- 2018-2019 Semester 2
- 2018-2019 Semester 1
- 2017-2018 Semester 2
- 2017-2018 Semester 1
- 2016-2017 Semester 2
- 2016-2017 Semester 1
- 2015-2016 Semester 1
- 2015-2016 Semester 2
- 2014-2015 Semester 2
- 2014-2015 Semester 1
- 2013-2014 Semester 2
- 2013-2014 Semester 1
- 2012-2013 Semester 2
- 2012-2013 Semester 1
- 2011-2012 Semester 2
- 2011-2012 Semester 1
- 2010-2011 Semester 2
- 2010-2011 Semester 1
- 2009-2010 Semester 2
- 2009-2010 Semester 1
- 2008-2009 Semester 2
- 2008-2009 Semester 1
- 2007-2008 Semester 2
- 2007-2008 Semester 1
- 2006-2007 Semester 2
- 2006-2007 Semester 1
- 2005-2006 Semester 2
- 2005-2006 Semester 1
- Contact
- Site Map
Robust Matrix Recovery through Nonconvex Optimization: Challenges and Promises
----------------------------------------------------------------------------------------------------
Department of Systems Engineering and Engineering Management
The Chinese University of Hong Kong
----------------------------------------------------------------------------------------------------
Date: Monday, April 22, 4:30 pm – 5:30 pm
Venue: ERB 513, The Chinese University of Hong Kong
Title: Robust Matrix Recovery through Nonconvex Optimization: Challenges and Promises
Speaker: Mr. Jianhao Ma, University of Michigan
Abstract:
Robust matrix recovery is a foundational problem in modern
machine learning, with applications in motion detection, face
recognition, and collaborative filtering. In this talk, we explore
overparameterized robust matrix recovery, where the true rank is
unknown, and investigate its algorithmic and landscape properties.
Firstly, we uncover that the true solutions are saddle points, rather
than local or global minima of the objective function. Surprisingly,
we demonstrate that a simple subgradient method with small
initialization converges to the true solution, achieving nearly
optimal sample complexity. Our findings offer insights into the
emerging paradigm of robust machine learning and emphasize the need
for novel theories in nonsmooth optimization.
Biography:
Jianhao Ma is currently a fourth-year Ph.D. student at the
Industrial and Operations Engineering Department of the University of
Michigan, Ann Arbor. His research lies broadly in robust machine
learning and nonconvex optimization. He received the Rackham
Predoctoral Fellowship from the University of Michigan. His papers
have won INFORMS JFIG Best Paper Award (second prize) and Katta Murty
Prize for Best Research Paper on Optimization (from IOE Department).
Everyone is welcome to attend the talk!
SEEM-5202 Website: http://seminar.se.cuhk.edu.hk
Email: seem5202@se.cuhk.edu.hk
Date:
Monday, April 22, 2024 - 16:30