Parameter Bounds Under Misspecified Models and Some Perspectives for Data Science and Learning
Christ D. Richmond, PhD
School of Electrical, Computer and Energy Engineering Arizona State University
Friday, April 13, 2018|
4:00pm - 5:00pm
Weiser Hall, Low Rise, Room 182
Add to Google Calendar
About the Event
Data models are the basis for many fundamental ideas and approaches in signal processing and statistical inference. These models, when correct, allow determination of fundamental limits on detection (binary hypothesis testing), classification/association (M-ary hypothesis testing), parameter estimation, data compression, and information rate transfer. Data models, however, are rarely specified correctly, as modeling errors at some level often persist; thus, diminishing to some degree the practical relevance and value of these classical performance limits. Recent advances in parameter estimation theory will be presented that identify fundamental bounds on estimation accuracy when data models are possibly misspecified. Some perspectives will be offered on the potential role such measures might play in the age of big data, as evidence suggests that modeling at some level continues to play an important role even with a deluge of data.
Christ D. Richmond is an Associate Professor in the School of Electrical, Computer and Energy Engineering at Arizona State University (ASU) where he leads the Signals, Inference, Information and Learning Group. His research interests include statistical signal processing, detection and parameter estimation theory, information theory, machine learning, radar/sonar, communications, and spectral sharing. Prior to joining ASU, he was a Senior Staff in the Advanced Sensor Techniques Group at the Massachusetts Institute of Technology (MIT) Lincoln Laboratory, and a Visiting Lecturer and Associate of the John A. Paulson School of Engineering and Applied Sciences at Harvard University. He received the Ph.D. degree in electrical engineering from MIT in 1996. He is the recipient of the Office of Naval Research Graduate Fellowship Award, the Alan Berman Research Publications Award, and the IEEE Signal Processing Society Young Author Best Paper Award in the area of Sensor Array and Multichannel (SAM) Signal Processing. He has served as the Technical Chairman of the Adaptive Sensor Array Processing Workshop at MIT Lincoln Laboratory, and served as a member the IEEE Technical Committee on SAM Signal Processing. He served as an Associate Editor for the IEEE Transactions on Signal Processing, and currently serves on the IEEE Warren D. White Award Selection Committee for The Art of Radar Engineering.
Contact: Alison Martin
Phone: (734) 615-8945
Faculty Sponsor: Al Hero
Open to: Public