|
Abstract:
|
We study the maximum score statistic to detect and estimate local signals in the form of change-points in the level, slope, or other property of a sequence of observations, and to segment the sequence when there appear to be multiple changes. We find that when observations are dependent, the change-points can lead to upwardly biased estimates of autocorrelations, resulting in a sometimes serious loss of power. Applications to temperature variations, atmospheric CO2 levels, disease incidence, and fluctuations in the size of animal populations, illustrate the general theory.
|
