Structure Learning Of BN Usine Improved Bacteria Foraging Optimization Algorithm

Bayesian Networks(BNs)is a graph-based network based on probabilistic reasoning to solve the problem of uncertainty and incompleteness.With the rise of artificial intelligence and data mining,the study of Bayesian networks has become a hotspot.The research of Bayesian Network Structure Learning,which is one of the NP-hard problems,has become the focus of research.Firstly,the paper expounds the origin and development of Bayesian network and the application of Bayesian network.Combined with a simple Bayesian network structure,it introduced the basic knowledge of Bayesian networks and Bayesian classifier knowledge,it also introduces the theoretical basis and research status of the bacterial algorithm.Secondly,according to the characteristics of Bayesian network,the bacterial algorithm has been improved accordingly.According to the traditional bacterial algorithm,A Bayesian network structure learning strategy based on improved bacterial foraging optimization algorithm is proposed for Bayesian network.The chemotaxis operator,the propagation operator and the migration operator in the traditional bacterial algorithm are improved.The adaptive theory was applied to the calculation of bacterial swimming step length and selection of breeding individuals.In the migration probability calculation of the migration operator,the roulette method in the genetic algorithm is introduced.On the basis of mutual information theory,a new stochastic evolutionary method of network structure is proposed,which replaces the random migration in traditional bacterial algorithm,and the improved bacterial algorithm is applied to Bayesian network structure learning.Finally,compare the network structure obtained by the algorithm with the greedy algorithm,GTT algorithm and K2 algorithm,the experiments show that the proposed algorithm can obtain better results in Bayesian network structure learning.