GMM EM Algorithem implementation
I Created a simple example for Bivariate GMM,which is an unsupervised learning algorithm,and a use case of EM algorithem. It can be used for clustering of data.
just run the file..
i wrote it in pycharm enviorment,python 3.7. so you may need to add some modules except for what i used- sklearn,numpy,random and scipy
you can run the training part in one piece, or run every function in separate.
common issue is that multivariate_normal.pdf scipy function can get nan or inf values and than it breaks, i generated for this an exception.if you are having trouble with that let me know.
Asael Bar Ilan you can contact me through e-mail-asaelbarilan@gmail.com
well few nice sources: 1.Expectation Maximization - Math and Pictures by Johannes Traa http://cal.cs.illinois.edu/~johannes/research/EM%20derivations.pdf-really explains the math behind very well 2. and this nice lecture too- http://www.ee.bgu.ac.il/~haimp/ml/lectures/lec2/lec2.pdf