Variational maximization-maximization of Bayesian mixture models and application to unsupervised image classification

• Main Author: Lim, Kart-Leong
• Other Authors: Wang Han
• Format: Theses and Dissertations
• Language: English
• Published: 2018
• Subjects: DRNTU::Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence
• Online Access: http://hdl.handle.net/10356/73199
• Institution: Nanyang Technological University

This thesis mainly propose variational inference for Bayesian mixture models and their applications to solve machine learning problems. The mixture models addressed are the Gaussian mixture model (GMM), Dirichlet process mixture (DPM), the sparse coding based Gaussian mixture model (sGMM) and the Fi...