Graphical user's interface.
Computation of the Independence Model, Homogeneous Markov Chains of any order, Hidden Markov Models, Double Chain Markov Models, and Mixture Transition Distribution models.
Hierarchical hidden models.
Simplified functions for the computation of the most common models.
Modeling of high-order Markov chains.
Use of covariates.
Handling of very large data files.
Different computation methods adapted to each model: Exact maximum-likelihood estimation, Expectation-Maximization (EM) algorithm, Best Improving Step algorithm for mixture models, Genetic Algorithm.
Algorithms can be combined in order to improve the estimation performances.
User's guide including theoretical notes on Markovian models and four tutorials.
Analysis of transition matrices: Identification of the type of each state, Decomposition of states into final classes, Computation of stable distributions, ...
Advanced tools for the analysis of hidden state sequences: Graphical display of hidden state sequences, Comparison of hidden state and observation sequences, Association measures between hidden states and observation sequences, Export of hidden state sequences, ...
Statistics for model comparison: Log-likelihood, Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC).
Computation of confidence interval for all parameters.
Analysis of the long-term behavior of a model through the successive powers of its transition matrix.
Simulation of data sequences from a given model.
Possibility to define theoretical Markovian models and to apply them on chosen data sets.
Easy management of results through the use of standard text files.
Diary file recording automatically all results obtained since the last opening of the software.
Many saving options for optimal management of model results.
Choice of the number of decimal places for results rounding.