Early Work (1992-2002)

 

 

Probabilistic Reasoning: Our most influential work in this area is an inference algorithm for Bayesian networks (BN) called variable elimination (Zhang and Poole 1994, 1996). It is the first inference algorithm for BN discussed in an popular BN textbook by Koller and Friedman and another popular AI textbook by Russell and Norvig, and the first inference algorithm for BN discussed in related courses offered at many universities. We have also proposed methods for exploiting causal independence and contextual independence in Bayesian network inference (Zhang and Poole 1996, 1999, and Poole and Zhang 2003) .

        N. L. Zhang and D. Poole (1996), Exploiting causal independence in Bayesian network inference,Journal of Artificial Intelligence Research, 5: 301-328. [Google citation count: 495]

        N. L. Zhang and D. Poole (1994), A simple approach to Bayesian network computations, in Proc. of the 10th Canadian Conference on Artificial Intelligence, Banff, Alberta, Canada, May 16-22. [Google citation count: 336]  (Featured at BN-wiki )

        D. Poole and Nevin L. Zhang (2003). Exploiting contextual independence in probabilistic inference. Journal of Artificial Intelligence Research, 18: 263-313.  [Google citation count: 122]

 

Decision-Theoretic Planning with POMDPs:  We have two notable results in this area: an exact algorithm called incremental pruning (IP) (Zhang and Liu 1997, Cassandra et al. 1997) and an approximate algorithm called pointed-based value iteration (PBVI) (Zhang and Zhang 2001).  IP is fundamental to the theory of POMDPs, while PBVI is a the key to make POMDPs practical. A large number of papers on PBVI were published subsequent to our work.

        Anthony Cassandra, Michael L. Littman, and N. L. Zhang (1997), Incremental Pruning: A Simple, Fast, Exact Algorithm for Partially Observable Markov Decision Processes in Proc. of the 13th Conference on Uncertainties in Artificial Intelligence. [Google citation count: 485]

        N. L. Zhang and W. Liu (1997), A model approximation scheme for planning in partially observable stochastic domains, Journal of Artificial Intelligence Research, 7: 199-230. [Google citation count: 85]

        N. L. Zhang and W. Liu (1996), Planning in stochastic domains: Problem characteristics and approximation.  Technical Report  HKUST-CS96-31  [Google citation count: 97

        Zhang, N.L. and Zhang, W. (2001) "Speeding up the Convergence of Value Iteration in Partially Observable Markov Decision Processes", Journal of Artificial Intelligence Research, 14: 29-51. [Google citation count: 149]

 

Decision under Uncertainty:  In this area, we proposed a general framework for representing and solving decision problems (Zhang et al 1994), and showed how general Bayesian network inference algorithm can be utilized to find optimal decisions in the framework (Zhang 1998) .

        N. L. Zhang (1998), Probabilistic Inference in Influence Diagrams, Computational Intelligence , 14(4):  475-497. [Google citation count: 117]

        N. L. Zhang R. Qi and D. Poole (1994) A computational theory of decision networks, International Journal of Approxi mate Reasoning, 1994, 11 (2): 83-158.  [Google citation count: 78]