Change Point Detection

For the at most one change point problem, we propose the use of a conceptor matrix to learn the characteristic dynamics of a specified training window in a time series.