Bayesian Inference of Mixture Model via Differential Evolution and Sampling

摘要：

Mixture model comprises a finite or infinite number of different distributional types of components and offers a much wider range of modeling possibilities than its components. In his paper, we present an approach for Bayesian inference of mixture model with Differential Evolution and Markov chain Monte Carlo(MCMC). Bayesian inference on Gaussian mixture model via Gibbs sampling and optimization with Differential Evolution MCMC are focuses of our work. The inference framework involves calculations of weight, mean and covariance corresponding to each component Experimental results show novel effect of our method.

关键词： Bayesian Inference Mixture Model Differential Evolution MCMC Gibbs sampling

作者: PengGuo Naixiang li

作者单位: School of Computer Science and Technology Tianjin University Tianjin China Department of Computer Sc Department of Computer Science and Information Engineering Tianjin Agricultural University Tianjin C

会议类型: 国际会议

会议名称: The 2nd IEEE International Conference on Advanced Computer Control(第二届先进计算机控制国际会议 ICACC 2010)

会议地点: 沈阳

会议语种:英文

页码: 504-508

在线出版日期: 2010-03-27（万方平台首次上网日期，不代表论文的发表时间）