Minimum distance classifier is a parametric classifier, because it is parameterized by the mean of the each class. It is special case of the Bayes classifier when the co-variance matrix is identity. Where as K-NN classifiers are non-parametric, eventhough they use a distance matrix, there is no class specific parameters. K-NN classifiers the unknown data based on the majority of the k-nearest training samples. The performance varies with the selection of 'K' neighbors.

It is parametric classifier