报告题目：Addressing the Computational Complexity Problems in Building a POMDP-based Intelligent Tutoring System

主讲人（职称，职务，工作单位）：王方矩，教授，加拿大圭尔夫大学

时间：2017年  10  月   18  日，10：30 - 11：20

地点：计算机科学与工程学院会议室F区113

Abstract    In building an intelligent tutoring system (ITS), the partially observable Markov decision process (POMDP) model provides useful tools for handling uncertainties, which are major challenges in achieving adaptive teaching. However, the POMDP model is very expensive. Typically the state space and solution space are both exponential. To build POMDP based ITSs for practical teaching, we need efficient techniques to address the computational complexity problems. In this talk, we will introduce the POMDP model and a POMDP framework for building ITSs, analyze the computational complexity problems, and present the techniques we developed to improve efficiency.