Causal State Modeller Toolbox

Allows users to infer minimal and optimal statistical predictors from time series data.

A reimplementation of the CSSR algorithm of Shalizi et al. Extends the input data from discrete to continuous clustered by way of a novel discretisation scheme. Allows users to infer optimally predictive hidden Markov models from their data, either supplied or generated by the code. Calculates the statistical complexity and other metrics of the data/inferred model. See help file for more details.