When does it make sense to try to learn a Mixture model from a data set? (Choose Multiple answers)

A. When the data set is multimodal, so it is unlikely that a simple distribution will fit the data well.

B. When we want to segment the data into different classes

C. When we want to be able to describe a complex data set in terms of a number of simple components

D. When our dataset is not linearly separable and we want to use the kernel-trick

E. None of the above.

F. All of the above.