STA 290 Seminar Series
Thursday, May 12th, 4:10pm, MSB 1147 (Colloquium Room)
Refreshments at 3:30pm in MSB 4110 (Statistics Lounge)
Speaker: Yajun Mei (Georgia Institute of Technology)
Title: “Scalable SUM-Shrinkage Schemes for Monitoring Large-Scale Data Streams”
Abstract: In the modern information age one often monitors large-scale data streams with the aim of offering the potential for early detection a ``trigger" event, e.g., biosurveillance, health care, security and environmental science. In this talk, we develop efficient scalable global monitoring schemes by combining parallel and distributed computing with the statistical tools from sequential change-point detection and shrinkage. In the first part of this talk, we provide a brief overview of the classical sequential change-point detection problem where one is monitoring a single data stream whose distribution may change at an unknown time. In the second part of the talk, we extend the classical sequential change-point detection algorithms to high-dimensional setting via shrinkage. Both asymptotic analysis and numerical simulations demonstrate the usefulness of shrinkage in the context of online monitoring large-scale data streams.
Bio: Dr. Mei received the Ph.D. degree in Mathematics with a minor in Electrical Engineering from the California Institute of Technology in 2003, and the B.S. degree in Mathematics from Peking University, P. R. China in 1996. He is currently an associate Professor in the H. Milton Stewart School of Industrial and Systems Engineering (ISyE) at the Georgia Institute of Technology. He was a Postdoctoral Researcher in biostatistics and biomathematics in the Fred Hutchinson Cancer Research Center, Seattle, WA, from 2003 to 2005. He received 2009 Abraham Wald Prize in Sequential Analysis and an NSF CAREER award in 2010. His research interests include change-point detection, sensor networks, sequential analysis, and their applications in engineering and biomedical sciences.