Change-point problems arise in a variety of experimental and mathematical sciences, as well as in engineering and health sciences. This rigorously researched text provides a comprehensive review of recent probabilistic methods for detecting various types of possible changes in the distribution of chronologically ordered observations. Further developing the already well-established theory of weighted approximations and weak convergence, the authors provide a thorough survey of parametric and non-parametric methods, regression and time series models together with sequential methods. All but the most basic models are carefully developed with detailed proofs, and illustrated by using a number of data sets. Contains a thorough survey of:

The Likelihood Approach

Non-Parametric Methods

Linear Models

Dependent Observations

This book is undoubtedly of interest to all probabilists and statisticians, experimental and health scientists, engineers, and essential for those working on quality control and surveillance problems.

Foreword by David Kendall