For processing backward reasoning and resolving the discrepancy between logic and probabihtv insequential decision problem with high order conditionals, we propose a backward Bavesian probabilistic logicreasoning approach.We propose a second-order backward Bavesian probabihstic logic reasoning approach, which combines CEAand Gibbs sampler to evaluate the quantitative value of second-order conditional probabihty.rnInprocessing backward reasoning in a decision probe usepartly change casual relations at first. And thenconditional event to denote casual relations forresolving the discrepancy between probabihtv and classiclogic, and transform higher-order conditional event tonormal events and corresponding joint events. By usingGibbs sampler, we obtain the normal events and jointevents in a stationary state. The quantitative value ofhigher-order conditional event is estimated, and we finish the backward reasoning.
展开▼
