Probabilistic clustering
Basic questions
what’s the probability that a point x is in cluster m?
what’s the shape of each cluster?
K-means doesn’t answer these questions
Probabilistic clustering (basic idea)
Treat each cluster as a Gaussian density function