Scalable robust hypothesis tests using graphical models

Divyanshu Vats, Vishal Monga, Umamahesh Srinivas, José M.F. Moura

    Research output: Chapter in Book/Report/Conference proceedingConference contribution

    1 Scopus citations

    Abstract

    Traditional binary hypothesis testing relies on the precise knowledge of the probability density of an observed random vector conditioned on each hypothesis. However, for many applications, these densities can only be approximated due to limited training data or dynamic changes affecting the observed signal. A classical approach to handle such scenarios of imprecise knowledge is via minimax robust hypothesis testing (RHT), where a test is designed to minimize the worst case performance for all models in the vicinity of the approximated imprecise density. Despite the promise of RHT for robust classification problems, its applications have remained rather limited because RHT in its native form does not scale gracefully with the dimension of the observed random vector. In this paper, we use approximations via probabilistic graphical models, in particular block-tree graphs, to enable computationally tractable algorithms for realizing RHT on high-dimensional data. We quantify the reductions in computational complexity. Experimental results on simulated data and a target recognition problem show minimal loss over a true RHT.

    Original languageEnglish (US)
    Title of host publication2011 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011 - Proceedings
    Pages3612-3615
    Number of pages4
    DOIs
    StatePublished - Aug 18 2011
    Event36th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011 - Prague, Czech Republic
    Duration: May 22 2011May 27 2011

    Other

    Other36th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2011
    Country/TerritoryCzech Republic
    CityPrague
    Period5/22/115/27/11

    All Science Journal Classification (ASJC) codes

    • Software
    • Signal Processing
    • Electrical and Electronic Engineering

    Fingerprint

    Dive into the research topics of 'Scalable robust hypothesis tests using graphical models'. Together they form a unique fingerprint.

    Cite this