Linear Algebra Seminar

Speaker: Xiaofei Wang (UAB)

Title: Latent Tree Model

Abstract: We talk about a tree-structure graphical probabilistic model where all the leaf nodes are observed while the internal nodes are latent, which is referred to as latent tree model. This model is gaining popularity in Machine Learning due to applications to speech recognition, computer vision, and bioinformatics. We will mainly discuss two topics of this model: structural learning and parameter estimation, and introduce some application of it on traditional Chinese medicine and text mining.