Page 7 - 第四届运筹青年论坛会议手册-0615

邀请报告 | 第四届中国运筹青年论坛 05

Multi-Dimensional Scaling and Modern Matrix Optimization

戚厚铎 英国南安普敦大学

Abstract: Multi-Dimensional Scaling (MDS) is deeply rooted in human perceptions of data and has

been developed and widely used in psychology. It is much relevant to today's big data science and

raises many questions that may be tackled from optimization. This talk gives a brief introduction of

the historical development of MDS. We then try to understand it from matrix optimization, aiming
to investigate whether the fast developing modern matrix theory and algorithms can lead to

fundamental contributions to the field of MDS. We demonstrate its power and challenges it faces

through the following applications: Dimensionality reduction in machine learning; Sensor network

localization; Molecular conformation; and Quadratic entropy in financial portfolio.

Biodata: Houduo Qi received the BSc in Statistics from Peking University in 1990, MSc in

Operational Research and Optimal Control from Qufu Normal University in 1993, and PhD in
Operational Research from Institute of Applied Mathematics, Chinese Academy of Sciences (CAS)

in 1996. He had been a postdoctoral research fellow in various institutions including Institute of

Computational Mathematics, CAS, The Hong Kong Polytechnic University, and The University of

New South Wales. In September 2004, he joined the University of Southampton as a lecturer in

Operational Research, rising to Professor and Chair of Optimization. His current research interest is

in Matrix Optimization with applications to finance and engineering sciences. He currently acts as

the Area Editor (Optimization) for Asia-Pacific Journal of Operational Research, Associate Editor

for Mathematical Programming Computation, and Journal of Operations Research Society of China.

From 2010, he has been a college member of Engineering and Physical Sciences Research Council,

UK. He is also Turing fellow at the Alan Turing Institute – UK’s national data science institute.