MODL, by Marc BOULLÉ
The Edge-ML algorithms of discretisation, grouping, extraction of sequential rules and supervised classification have been developped by Alexis BONDU from the MODL approch concived and published py Marc BOULLE – Orange Labs Lannion.
MODL is a Bayesian model selection approach. Its singularity comes from a mathemathical formalism at the crossroad of the Machine Learning and the Information Theory. This very smart approach, without regularization parameters to be ajusted by a grid-search, avoids over-fitting. MODL provides robust models based on a rigorous mathematical formalism. These models reach an excellent compromise between accuracy, robustness and rapidity.
For further information :
 M. Boullé. MODL: a Bayes optimal discretization method for continuous attributes. Machine Learning, 65(1):131-165, 2006.
 M. Boullé. A Grouping Method for Categorical Attributes Having Very Large Number of Values. ICDM, LNAI, Volume 3587, 2005.
 M. Boullé. Compression-Based Averaging of Selective Naive Bayes Classifiers. Journal of Machine Learning Research, 8:1659-1685, 2007.
 M. E. Egho, D. Gay, N. Voisine, M. Boullé, F. Clérot. A Parameter-Free Approach for Mining Robust Sequential Classification Rules. ICDM 2015.