Uncertain Knowledge Representation
Richard E. Neapolitan, Xia Jiang · 2018
Pearl (1986) conjectured that uncertain knowledge is structured with causal edges between propositions. This chapter includes the variables for age and sex in the network when the age and sex of the card holder has nothing to do with whether the card has been stolen. The DAG and the conditional distributions together constitute a Bayesian network. The chapter discusses why a causal DAG should satisfy the Markov condition with the probability distribution of the variables in the DAG. It argues that a causal DAG often satisfies the Markov condition with the joint probability distribution of the random variables in the DAG. The chapter presents some examples illustrating how the conditional independencies entailed by the Markov condition can be exploited to accomplish inference in a Bayesian network. It illustrates inference in a Bayesian network using Netica.