This MATLAB code is based on the paper 'MAXIMUM A-POSTERIORI ESTIMATION OF MISSING SAMPLES WITH CONTINUITY
CONSTRAINT IN ELECTROMAGNETIC ARTICULOGRAPHY DATA' 

First, parameters for GMM is learned from the training data which does not contain any missing samples using the function gmm_parameters_learn.

Missing samples in the test data are estimated using the function missing_samples_estimate


For illustration of the missing samples estimation, please follow the 'example_progam', the sample_missing_indices can be changed to change the missing data locations.