Performance Evaluation of Some Change Detection and Data Segmentation Algorithms

Abstract: The problem of change detection and data segmentation has received considerable attention during the last two decade in a research context and appears to be the central issue in various application areas. The following techniques are investigated in the paper for their performance evaluation: filtering techniques with a whiteness test, techniques based on sliding windows and distance measures and maximum likelihood techniques for data segmentation. The used model will be the simplest extension of linear regression models to data with abruptly changing properties, or piecewise linearizations of non-linear models. Finally, some Monte-Carlo simulations for change detection and data segmentation are presented, to evaluate the performance of these algorithms in a number of cases.