GitHub - stochasticresearch/mutual-information-ICA: Matlab scripts for implementing a mutual-information-based algorithm for independent component analysis. The algorithm works by iteratively rotating pairs of principle components until an optimum is found. The cost function is based on the summed negentropy of the discovered components, which is closely relative to the component mutual information.

Branches Tags

Name		Name	Last commit message	Last commit date
Latest commit History 9 Commits
compute_signal_similarities.m		compute_signal_similarities.m
entropy.m		entropy.m
histogram.m		histogram.m
min_mi_estimator_ICA.m		min_mi_estimator_ICA.m
minimize_mutual_information_via_rotation.m		minimize_mutual_information_via_rotation.m
minimize_via_rotation.m		minimize_via_rotation.m
mutual_information_ICA.m		mutual_information_ICA.m
random_rotation_matrix.m		random_rotation_matrix.m
test_mi_estimators.m		test_mi_estimators.m
test_mi_ica.m		test_mi_ica.m

About

Matlab scripts for implementing a mutual-information-based algorithm for independent component analysis. The algorithm works by iteratively rotating pairs of principle components until an optimum is found. The cost function is based on the summed negentropy of the discovered components, which is closely relative to the component mutual informati…