Scripts for (curve) clustering with Regression Mixtures (polynomial, spline, B-spline, with mixed random effects) and the robust EM algorithm. The algorithm simultaneously estimates the model parameters, and the number of mixture components, by a regularized likelihood estimation approach. Please cite the following papers and this code when using it:

• F. Chamroukhi. Unsupervised learning of regression mixture models with unknown number of components. Journal of Statistical Computation and Simulation Taylor & Francis Online., Vol. 86, pages:2308-2334, Nov, 2016
• F. Chamroukhi. Robust EM algorithm for model-based curve clustering., pages:1-8, August, 2013, Proceedings of the International Joint Conference on Neural Networks (IJCNN), IEEE

(c) by F. Chamroukhi