Bayesian model-averaging in unsupervised learning from microarray data

摘要: Unsupervised identification of patterns in microarray data has been a productive approach to uncovering relationships between genes and the biological process which they are involved. Traditional model-based clustering approaches as well some recently developed mining for integrating genomic functional rely on one's ability determine correct number clusters or modules data. In this paper we demonstrate that performance such methods general can be significantly improved by accounting uncertainties inherent identifying optimal We Bayesian averaging via infinite mixture model offers more robust than traditional finite is determined using Information Criterion. This improvement demonstrated through simulation study analysis relatively large dataset. Finally, describe novel heuristic modification Gibbs sampler used fit mode effectively deals with issues slow mixing.