ZHAO, HE Structured Bayesian Latent Factor Models with Meta-data In the era of big data, huge amounts of data are being generated from the internet, social networks, phone apps, and so on, which creates high demand for powerful and efficient data analysis techniques. In areas such as collaborative filtering, text analysis, graph analysis, and bioinformatics, a large proportion of such data can be formulated into discrete matrices. This research focuses on developing structured Bayesian latent factor models with meta-data for analysing discrete data in the above areas. Compared with state-of-the-art methods, the proposed approaches have achieved not only better modelling performance and efficiency, but also preferable interpretability for intuitively understanding those data. Machine Learning;Bayesian Analysis;Text Analysis;Graph Analysis;Bayesian Nonparametrics;Probabilistic Graphical Model;Matrix Factorisation;Latent Variable Model;Topic Model;Meta-data;Knowledge Representation and Machine Learning;Applied Statistics;Applied Computer Science 2019-07-08
    https://bridges.monash.edu/articles/thesis/Structured_Bayesian_Latent_Factor_Models_with_Meta-data/8813471
10.26180/5d23c9a348a83