EM Mixture of Probabilistic PCA, Python3 (ongoing)

Implementation of the paper : Michael E. Tipping and Christopher M. Bishop, Mixtures of Probabilistic Principal Component Analysers

Implementation:

• This implementation was based at a first stage on Michel Blancard's repo https://github.com/michelbl/MPPCA
• Added the base Probabilistic PCA model with classic methods : fit, transform, and score
• For a MPPCA, Blancard's repo can also be used which provides also a python notebook to visualize results.