Error exponents for composite hypothesis testing of Markov forest distributions

Error exponents for composite hypothesis testing of Markov forest distributions

Author: Tan, Vincent Y. F.; Anandkumar, Animashree; Willsky, Alan S.

Citable URI: http://hdl.handle.net/1721.1/73578

Department: Massachusetts Institute of Technology. Laboratory for Information and Decision Systems; Massachusetts Institute of Technology. Dept. of Electrical Engineering and Computer Science

Publisher: Institute of Electrical and Electronics Engineers (IEEE)

Date Issued: 2010-07

Abstract:

The problem of composite binary hypothesis testing of Markov forest (or tree) distributions is considered. The worst-case type-II error exponent is derived under the Neyman-Pearson formulation. Under simple null hypothesis, the error exponent is derived in closed-form and is characterized in terms of the so-called bottleneck edge of the forest distribution. The least favorable distribution for detection is shown to be Markov on the second-best max-weight spanning tree with mutual information edge weights. A necessary and sufficient condition to have positive error exponent is derived.

URI: http://hdl.handle.net/1721.1/73578

ISBN: 978-1-4244-7891-0
978-1-4244-7890-3

Citation: Tan, Vincent Y. F., Animashree Anandkumar, and Alan S. Willsky. “Error Exponents for Composite Hypothesis Testing of Markov Forest Distributions.” IEEE International Symposium on Information Theory Proceedings (ISIT), 2010. 1613–1617. © Copyright 2010 IEEE

Version: Final published version

Terms of Use: Article is made available in accordance with the publisher's policy and may be subject to US copyright law. Please refer to the publisher's site for terms of use.

Published As: http://dx.doi.org/10.1109/ISIT.2010.5513399

Journal: Proceedings of the IEEE International Symposium on Information Theory Proceedings (ISIT), 2010

Show full item metadata