Stanford ml Public Lesson notes 12--k-means, mixed Gaussian distributions, EM algorithm

PDF documents for the Open Class series have been uploaded to Csdn resources, please click here to download.

This article corresponds to the 12th video of the Stanford ML Public course, and the 12th video is not very relevant to the previous one, opening a new topic-unsupervised learning. The main contents include the K-means clustering (K-means) algorithm in unsupervised learning, the mixed Gaussian distribution model (Mixture of Gaussians, MoG), the EM algorithm for solving the MoG model, and the generalized form of EM, before the generalized form of EM, there is a small knowledge point, That is, the Jensen inequality (Jensen ' s inequality).