I have a Gaussian model, I would like to classify new experimental data. Does Mahalanobis distance approach is a good starting point to evaluate the probability distribution?
Or is there any other approach would provide better results?
Expert advice is appreciated..:)