In time-series data, there are some abrupt changes which can indicate a statistical alteration of upcoming samples of data. They are called as change-points. The time point which these abrupt changes occur can give useful information according to the type of data. In order to detect these abrupt changes, change-point detection methods are utilized. As online change-point detection methods, Area Under the Receiver Operating Characteristics Curve (AUROC), 2-sample Kolmogorov-Smirnov Test and Bayesian Online Change-Point Detection are investigated. The performances are experimented using simulated time-series data including two different experiments; Mean Test and Variance Test.
