A Sequential Framework for Composite Hypothesis Testing

Shahar Bar, Joseph Tabrikian

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

This paper addresses the problem of classification via composite sequential hypothesis testing. We focus on two possible schemes for the hypotheses: non-nested and nested. For the first case, we present the generalized sequential probability ratio test (GSPRT) and provide an analysis of its asymptotic optimality. Yet, for the nested case, this algorithm is shown to be inconsistent. Consequently, an alternative solution is derived based on Bayesian considerations, similar to the ones used for the Bayesian information criterion and asymptotic maximum a posteriori probability criterion for model order selection. The proposed test, named penalized GSPRT (PGSPRT), is based on restraining the exponential growth of the GSPRT with respect to the sequential probability ratio test. Furthermore, the commonly used performance measure for sequential tests, known as the average sample number, is evaluated for the PGSPRT under each of the hypotheses. Simulations are carried out to compare the performance measures of the proposed algorithms for two nested model order selection problems.

Original languageEnglish
Article number8444760
Pages (from-to)5484-5499
Number of pages16
JournalIEEE Transactions on Signal Processing
Volume66
Issue number20
DOIs
StatePublished - 15 Oct 2018

Keywords

  • Sequential hypothesis testing (SHT)
  • average sample number (ASN)
  • composite hypothesis testing
  • generalized sequential probability ratio test (GSPRT)
  • maximum likelihood

Fingerprint

Dive into the research topics of 'A Sequential Framework for Composite Hypothesis Testing'. Together they form a unique fingerprint.

Cite this