Discrete Dynamic Bayesian Network Threat Assessment Method Based on Cloud Parameter Learning

Aiming at the problem that the data is easily divided inaccurately at the conceptual junction after the discretization of the target continuous threat variable, a Discretization Dynamic Bayesian Network (DDBN)parameter learning algorithm based on cloud model is proposed. The cloud model is adopted to realize softening points from continuous quantitative data to qualitative concepts, to obtain the soft probability that the data belongs to the qualitative concept; On this basis, the state concept is expanded and replicated in proportion to probability, and multiple sets of hard evidence samples are obtained, and then the parameter learning is performed; the learned parameters are used for threat reasoning, and the posterior probability that the target belongs to each threat level is obtained. Threat ranking is carried out by introducing utility theory and the simulation results verify the rationality of the target threat assessment under the condition of the small sample data missing.