We propose a novel framework for performing quan-titative Bayesian inference based on qualitative knowledge. Here, we focus on the treatment in case of inconsistent qualitative knowledge. A hi-erarchical Bayesian model is proposed for integrat-ing inconsistent qualitative knowledge by calculat-ing a prior belief distribution based on a vector of knowledge features. Each inconsistent knowledge component uniquely defines a model class in the hyperspace. A set of constraints within each class is generated to describe the uncertainty in ground Bayesian model space. Quantitative Bayesian in-ference is approximated by model averaging with Monte Carlo methods. Our method is tested on ASIA network and results suggest that it enables reasonable quantitative Bayesian inference from a set of inconsistent qualitative knowledge.
展开▼
