Publication Date: July 31, 2009 ISBN-10: 0262013193 ISBN-13 :978-0262013192 Edition: 1
Most tasks require a person or an automated system – providing information on the basis of concluded. The framework of probabilistic graphical model, in this book, provides a generic way to accomplish this task. The method is based on the model, the interpretation of the model construction and manipulation of the inference algorithm. These models can also be used in the method of automatically learning from data manually, it is difficult to build a model, or even impossible. Because uncertainty is inevitable, most real-world applications, this book focuses on probabilistic model, explicit uncertainty and provide more faithful to the reality of the model, these models. Probabilistic graphical model, and discuss a variety of models, covering Bayesian networks, undirected Markov networks, discrete and continuous models, extended to deal with dynamical systems and relational data. For each type of model, three basic building blocks: the textual description, reasoning and learning, the basic concepts and advanced technology. Finally, this book argues, causal reasoning and decision-making under conditions of uncertainty of the proposed framework. The main content of each chapter provides a detailed technical development of the main ideas. Most chapters also include boxes and other material: skill boxes, which describes the technical case study boxes, which discusses the relevant experience, the method described in the text, including computer vision, robotics, natural language understanding and computational biological applications; concept boxes, and now the important concepts from a chapter in the material. Teachers (readers) in various combinations in the group of chapters from the core theme of the more technologically advanced materials to meet their specific needs.
