Constructive Computations of Applied Stochastic Models
主 题: Constructive Computations of Applied Stochastic Models
报告人: 李泉林 博士 (清华大学工业工程系)
时 间: 2008-12-08 下午 3:00 - 4:00
地 点: 理科一号楼 1418
In the talk, we indicate how to use Markov processes to study some practical applied systems such as computer networks, manufacturing systems, transportation networks and economic systems. Based on this, we summarize the main block-structured Markov chains, Markov decision processes and stochastic game models.
To analyze these above practical systems, we have developed a numerical method by means of the so-called UL-type and LU-type RG-factorizations in the past ten years. We show that the RG-factorizations play an important role in the study of stochastic models. Specifically, the RG-factorization method has formed a complete computational framework for large-scale or complicated stochastic systems.
We review the main results of Markov chains or stochastic models under the computational framework based on the RG-factorizations, and provide several promising problems in the future research.
