[sieisteinmodel]

The problem is that this guy has no clue.

- A naive Bayes classifier is not a Bayesian method. It's just using Bayes rule, while Bayesian methods marginalize out the parameters.

- Regression and classification are different things. And, ironically, logistic regression is a classifier.

- Neither Linear Discriminant Analysis nor radial basis functions are kernel methods.

- EM is not a clustering method, it is an optimization algorithm for probabilistic models of a specific structure.

- Learning vector quantization is not a neural net.

- ...