bisecting Kmeans
The main idea of the bisecting Kmeans algorithm is: First, all the points as a cluster, then divide the cluster into two, then select the cluster that can minimize the clustering cost function (that is, the sum of squared error) is divided into two clusters to proceed, until the number of clusters equals the number of users given by K.
Gaussian Mixture Model
The so-called mixed Gaussian model refers to estimating the probability density distribution of a sample, and the estimated model is the sum of several Gaussian models (specifically, several to be established before the model is trained). Each Gaussian model represents a class (a cluster). When the data in the sample is projected on several Gaussian models, the probabilities on each class are obtained respectively. Then we can select the class with the most probability as the result of the verdict.
latent Dirichlet Allocation
The idea of the LDA thematic model is to abstract the content of a document into multiple topics, each with its own words, and each document is given in the form of a probability distribution.
Clustering algorithm related