Scalable pseudo-likelihood estimation in hybrid random fields

Learning probabilistic graphical models from high-dimensional datasets is a computationally challenging task. In many interesting applications, the domain dimensionality is such as to prevent state-of-the-art statistical learning techniques from delivering accurate models in reasonable time. We introduce a hybrid random field (HRF) model for pseudo-likelihood estimation in high-dimensional domains. A theoretical analysis proves that the class of pseudo-likelihood distributions representable by HRFs strictly includes the class of joint probability distributions representable by Bayesian networks (BNs). In order to learn the structure of HRFs from data, we develop the Markov Blanket Merging (MBM) algorithm. Theoretical and experimental evidence shows that MBM scales up very well to high-dimensional datasets. As compared to other statistical learning techniques, MBM delivers accurate results in a number of link prediction tasks, while achieving also significant improvements in terms of c...

• Type: Conference
• Year: 2009
• Where: KDD
• Authors: Antonino Freno, Edmondo Trentin, Marco Gori
• Keywords: Bni Contains Xi, Data Mining, KDD 2009, Random Variables, Statistical Learning Techniques