The conditional independence of naive Bayes classifier was also known as "naive Bayesian hypothesis"

that limited the application range and classification accuracy of algorithm.Concerning to this problem

the improved whale optimization algorithm（which shortened as IWOA） was used to optimize the naive Bayesian classifier.The IWOA used taboo search mechanism to avoid the misunderstanding of falling local optimization.In order to weaken the independence assumption of naive Bayes classifier

the IWOA automatically searched the global weights of the attributes of the classifier

so as to improve the operation accuracy of the weighted Bayesian classifier.The experimental result shows that comparing with the traditional Bayes classification

this proposed algorithm is more accurate in the paper.