✍ Scribed by A. C. Liew; P. K. Dash
No coin nor oath required. For personal study only.
, H , 2 , H )\ at some unknown time. We desire to detect that a change has occurred and to isolate or identify the type H J , l" 1, 2 , K!1, of this change. Chapter 7 is devoted to this problem. The first section discusses the Armitage SPRT for the statistical multiple-hypotheses testing problem. Next a kind of multiple-hypotheses CUSUM algorithm is proposed and its extension to the case of auxiliary input is briefly described. Although no proofs of optimality or analytical performances of the proposed algorithm are given, the author's idea is closely related to an asymptotically optimal solution which has appeared since the publication of the book.
In conclusion, this is a well-written book. It should be of special interest for readers with some statistical and signal-processing backgrounds. Although some problems have been treated at the heuristic level, the book provides practising researchers with useful methods for the design of change detection algorithms. Researchers interested in the theoretical development of the area will find this book a rich source of ideas.
📜 SIMILAR VOLUMES