2015/04/28 by Xin Yuan, Ricardo Henao, Yuan, Xin +7
Biochemistry, Genetics and Molecular Biology · Computer Science · Economics, Econometrics and Finance · #Bayesian Methods and Mixture Models #FOS: Computer and information sciences #Financial Risk and Volatility Modeling #Gaussian Processes and Bayesian Inference #Gene expression and cancer classification #Machine Learning (stat.ML)
paper · pdf · doi:10.48550/arxiv.1504.07468
openalex publication_date 2015/04/28 · openalex created_date 2019/06/27 · openalex updated_date 2026/07/28
We consider the problem of discriminative factor analysis for data that are\nin general non-Gaussian. A Bayesian model based on the ranks of the data is\nproposed. We first introduce a new em max-margin version of the\nrank-likelihood. A discriminative factor model is then developed, integrating\nthe max-margin rank-likelihood and (linear) Bayesian support vector machines,\nwhich are also built on the max-margin principle. The discriminative factor\nmodel is further extended to the em nonlinear case through mixtures of local\nlinear classifiers, via Dirichlet processes. Fully local conjugacy of the model\nyields efficient inference with both Markov Chain Monte Carlo and variational\nBayes approaches. Extensive experiments on benchmark and real data demonstrate\nsuperior performance of the proposed model and its potential for applications\nin computational biology.\n