RMSS
Software:5982120
Robust Multi-Model Subset Selection
Anthony Christidis, Gabriela Cohen-Freue
Last update: 14 September 2023
Software version identifier: 1.0.0, 1.1.0, 1.1.1
Copyright license: GNU General Public License, version 3.0, GNU General Public License, version 2.0
Efficient algorithms for generating ensembles of robust, sparse and diverse models via robust multi-model subset selection (RMSS). The robust ensembles are generated by minimizing the sum of the least trimmed square loss of the models in the ensembles under constraints for the size of the models and the sharing of the predictors. Tuning parameters for the robustness, sparsity and diversity of the robust ensemble are selected by cross-validation.
