The Application Of ID3 Optimization Algorithm In The Problem Of Hotel Rooms' Renewal Tuition

With the development of database technology,the technology of data mining classification has also been rapid development,the methods of decision tree in classification are often used in many industry fields.As one of the most influential decision tree algorithm,ID3 algorithm and has been used widely.However,ID3 algorithm also has some shortcomings.For the problem of multivalued bias and the formula cost a lot of time in ID3 algorithm,this paper optimizes the ID3 algorithm by improving the ID3 algorithm and simplifying the formula.The main works are as follows:(1)The ID3 optimization scheme based on correction function.In the ID3 optimization scheme based on the correction function,the method of overcome multivalued bias is: correct sub tuples desired information by introducing the definition of correction function.The multivalued bias analysis shows that the improved ID3 algorithm can overcome the problem of multivalued bias.The method of overcome ID3 algorithm formula cost a lot of time is:simplify information gain formula by using the equivalent infinite small,making simplified information gain formula has eliminated the logarithm operation.Finally,combine the work of improved algorithm and simplify information gain formula,and put forward the decision tree optimization algorithm based on correction function.The experimental results show that compared with the ID3 algorithm,in the two aspects of the classification accuracy and operation time,the decision tree optimization algorithm based on correction function is better.(2)The ID3 optimization scheme based on correlation coefficient.In the ID3 optimization scheme based on the correction function,the method of overcome multivalued bias is:get a new correlation coefficient by making the correlation coefficient modified into applicable to the field of decision tree.Correct sub tuples desired information by introducing the definition of correlation coefficient.The multivalued bias analysis shows that the improved ID3 algorithm can overcome the problem of multivalued bias.The method of overcome ID3 algorithm formula cost a lot of time is:simplify information gain formula by using the Maclaurin formula,making simplified information gain formula has eliminated the logarithm operation.Finally,combine the work of improved algorithm and simplify information gain formula,and put forward the decision tree optimization algorithm based on correction coefficient.The experimental results show that compared with the ID3 algorithm,in the two aspects of the classification accuracy and operation time,the decision tree optimization algorithm based on correction coefficient is better.compared(3)with the decision tree optimization algorithm based on correction function,the decision tree optimization algorithm based on correction coefficient has higher classification accuracy and cost more operation time.Finally,solve the problem of hotel rooms renewal by using two kinds of ID3 optimization algorithm,and achieve optimization algorithm by using C + + language in Visual Studio2013.Make a data set by combining the user information and suggestion,create decision tree with ID3 algorithm and two optimization algorithm,and extract the rules.Through compare and analyse the result of the decision tree,it shows that two kinds of ID3 optimization algorithm have superiority.