|
A priori inequalities in sequential hypothesis testing and change-point detection problems
by
Boris S. Darkhovsky
Institute for Systems Analysis RAS, Moscow, Russia
Coauthors: Boris E. Brodsky (Central Institute for Mathematics and Economics RAS)
New criteria of effectiveness in multi-hypothesis sequential testing and CPD problems are proposed. These criteria are based upon a priori lower bounds for the ratio (in an appropriate scale) of the average time before the true decision to the average time before the false decision (in the least favorable situation of testing or detection). These lower bounds hold for a wide class of stopping times for independent observations. Methodological ties between new performance indicators and traditional criteria of effectiveness in sequential analysis and CPD problems are considered in detail.
Date received: February 19, 2007
Copyright © 2007 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # cauc-19.