FEMDA robust classification

Update 2022: our paper has just been accepted at ICASSP 2022 - the International Conference on Acoustics, Speech, & Signal Processing. ICASSP is the IEEE Signal Processing Society’s flagship conference on signal processing.

Cite: P. Houdouin, A. Wang, M. Jonckheere and F. Pascal, “Robust Classification with Flexible Discriminant Analysis in Heterogeneous Data,” ICASSP 2022 - 2022 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), 2022, pp. 5717-5721, doi: 10.1109/ICASSP43922.2022.9747576.

Output of my research at CentraleSupélec, during my 5-weeks internship in the Laboratoire des signaux et systèmes. Flexible EM-Inspired Discriminant Analysis is a robust supervised classification algorithm that performs well in heterogeneous or noisy and contaminated datasets. femda is implemented as a scikit-learn estimator and is available here.