A Method for Predicting Major Events Based on Bayesian Network
Aiming at the prediction needs of major events such as regional conflicts,machine learn-ing is introduced into the research field of major events,and the relationship between predicted events and influencing factors is expressed by using Bayesian network structure,and the probability of major events is obtained through the analysis of causal networks.In order to better match the semantic character-istics of data,regular expressions are used as matching rules.Considering the timeliness of the data,the time decay function is introduced and fuzzy random variables are used to describe the node state to improve the accuracy of the prediction results.This method can efficiently,accurately and automatically predict the occurrence probability of major events,and can provide certain auxiliary support for making strategic decisions in advance.
万方数据
维普
