GMM based Expectation Maximization Algorithm.zip

The code consist of the implementation of model based technique for data labelling or clustering.Each element of data is considered as random variable whose probability distribution function is Gaussian.In order to estimate model parameters EM is used to maximize the log likelihood of given data set.
Gaussian parameters obtained using EM algorithm are further used to calculate probabilities of data values for every class.And finally on the basis of maximum probability data values are labeled into classes.