Turning-Point Detection Toolbox for Matlab

This Matlab code provides a Graphical User Interface for the sequential detection of turning points in time series. It implements various statistical methods, both non-parametric (as explonential smoothing), and parametric (as Markov Switching). Smothing coefficients and system parameters are selected by automatic (data-driven) methods, based on maximun likelihood and gain optimization. The toolbox can process non-stationary time series with stochastic cycle which are present in Economics, Engeeniring, Earth Science and Medicine, for purposes of monitoring, surveillance, forecasting and control.