A probabilistic model of data within each class is an example of generative classification.
Generative models, in the context of machine learning, are a kind of statistical model that determines the probability distribution of the input data. These models operate differently from discriminative models, which instead differentiate between classes of input.
More specifically, in a generative classification model, the class membership is modeled as a random variable which is dependent on the observation values. We start by modeling the distribution of the data within each class separately. This involves understanding the factors which could generate the respective class. Then using Bayes' rule, we turn around these models to predict the class given a new observation.
The key here is that the generative approach models the distribution of individual classes. So, the correct answer is option B, Generative classification.