The repository contains two Matlab functions: crossvalidated_lasso.m and confusion_matrix.m, and an example script describing how to use them: crossvalidated_lasso_example.m The crossvalidated_lasso function implements LASSO regression based (sparse) classification with hyper-parameter learning based on nested cross-validation. The cross-validation schemes and the parameters of the LASSO regression can be defined by the user. The confusion_matrix function implements function accuracy, averaged F-measure and confusion matrix calculations for given predictions (0 or 1) and labels (0 or 1).
RESEARCH PAPERS USING THIS SOFTWARE:
- Meszlényi, Regina, Ladislav Peska, Viktor Gál, Zoltán Vidnyánszky, and Krisztian Buza. 2016a. ‘A Model for Classification Based on the Functional Connectivity Pattern Dynamics of the Brain’. Proceedings of the Third European Network Intelligence Conference (ENIC), IEEE, 2016.
- Meszlényi, Regina, Ladislav Peska, Viktor Gál, Zoltán Vidnyánszky, and Krisztian Buza. 2016b. ‘Classification of fMRI Data Using Dynamic Time Warping Based Functional Connectivity Analysis’. Proceedings of the 24th Signal Processing Conference (EUSIPCO),2016.