基于模型聚类算法的matlab实现 This does the entire MB Clustering given a set of data. It only does the 4 basic models, unequal-unknown priors. It returns the BESTMODEL based on the highest BIC.-model-based clustering algorithm to achieve the Matlab This do

This demo nstrates the use of the reversible jump MCMC simulated annealing for neural networks. This algorithm enables us to maximise the joint posterior distribution of the network parameters and the number of basis function. It performs a global se

he algorithm is equivalent to Infomax by Bell and Sejnowski 1995 [1] using a maximum likelihood formulation. No noise is assumed and the number of observations must equal the number of sources. The BFGS method [2] is used for optimization. The