Imperial College London > Talks@ee.imperial > Featured talks > Machine Learning; Directly Estimable Information Divergence Measures with applications

Machine Learning; Directly Estimable Information Divergence Measures with applications

Add to your list(s) Download to your calendar using vCal

  • UserAlan Wisler, PhD Candidate at Arizona State University & Professor Andreas Spanias, Director of SenSIP Center, Arizona State University
  • ClockFriday 06 November 2015, 15:00-16:00
  • HouseGabor Seminar Room, 611, level 6, EEE Dept. .

If you have a question about this talk, please contact Charlotte B Grady.

Collaborative Research with:
  • Visar Berisha, Professor ECEE and SHS , ASU
  • Al Hero, ECE , University of Michigan
  • Andreas Spanias, SenSIP Center, ECEE , ASU

Abstract The seminar will start with a short introduction of the SenSIP center by Prof. Spanias. It will then continue with a seminar on machine learning and its applications. Information divergence functions play a critical role in statistics and information theory. In this presentation we first present a new f-divergence measure that can be used to provide improved bounds on the minimum probability of error. Furthermore this divergence measure can be estimated directly from data – we present an asymptotically consistent estimator of the divergence measure that does not require density estimates of the two distributions. Following, we also outline a formulaic design process for how to construct data-driven estimators for existing divergence functions and how to design your own divergence with custom properties. Throughout the presentation we will complement the theoretical results with empirical results from various speech applications.

Biographies: Alan completed both his BS and MS degrees in electrical engineering at the University of Texas at Dallas. Since 2013 Alan has been pursuing his PhD at Arizona State University under advisers Dr. Visar Berisha and Dr. Spanias. His research is primarily focused on the study and development of empirically estimable algorithm-agnostic performance bounds and their applications to statistical learning challenges in speech processing.

Andreas Spanias is Professor in the School of Electrical, Computer, and Energy Engineering at Arizona State University (ASU). He is also the director of the Sensor Signal and Information Processing (SenSIP) center and the founder of the SenSIP industry consortium (now an NSF I /UCRC site). His research interests are in the areas of adaptive signal processing, speech processing, and sensor systems. He served as Associate Editor of the IEEE Transactions on Signal Processing and as General Co-chair of IEEE ICASSP -99. He also served as the IEEE Signal Processing Vice-President for Conferences. Andreas Spanias is co-recipient of the 2002 IEEE Donald G. Fink paper prize award and was elected Fellow of the IEEE in 2003. He served as Distinguished lecturer for the IEEE Signal processing society in 2004.

This talk is part of the Featured talks series.

Tell a friend about this talk:

This talk is included in these lists:

Note that ex-directory lists are not shown.

 

Changes to Talks@imperial | Privacy and Publicity