Font Size: a A A

The Association Rule Data Mining & Application Research Based On One Improved Genetic Algorithms

Posted on:2017-04-03Degree:MasterType:Thesis
Country:ChinaCandidate:Q WangFull Text:PDF
GTID:2348330488490499Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
With the rapid development of electronic commerce and the Internet,information technology and digitization become characteristic of the era.In various industries the database store a lot of data,these data contain a lot of valuable information,while association rule mining excavate hidden and valuable relationships from large amounts of data information,and provide a strong basis for making right decision.Today,in many areas association rules mining for decision support has played an important role,mining strong association rules from transaction databases,help decision makers make decisions.While classical association rule mining algorithm-Apriori algorithm has low efficiency and long-running time in the vast amount of information.In order to compensate for the lack Apriori algorithm,this paper introduces the global search algorithm-genetic algorithm,and improve the genetic algorithm,and then puts forward an association rule mining algorithm based on improved genetic algorithm.Test the superiority of the algorithm by comparing the experiment,finally the algorithm is applied to a business website transactions data and evaluation data,verify the algorithm practicality.The main work is as follows:(1)Association rule mining: In-depth study of the association rule mining algorithm,elaborated on association rule mining classical algorithm-Apriori algorithms,analyze the advantages and disadvantages of common association rules mining algorithm,and raise an association rule mining using improved genetic algorithms.(2)Genetic algorithms: comprehensive introduction the basic concepts and ideas of genetic algorithms,comparative analysis of various coding methods,constructors of fitness function,designs of genetic operator in detail,and then make a detailed explanation for the step process of an improved genetic algorithm.(3)An association rule mining based on improved genetic algorithm: Design an association rule mining based on improved genetic algorithm,and make comparative tests for improved genetic algorithm,genetic algorithm,Apriori algorithm using MATLAB.Verify the superiority of the improved genetic algorithm for Mining Association Rules in this thesis from the number of association rules mining,processing running time of processing different data sets and running time for different support thresholds.(4)The application of an association rule mining based on improved genetic algorithm: A association rule Based on Improved Genetic Algorithm Mining applie to a business website,digging business website transactions found relationships between commodities;mining product evaluation data found that different types of merchandise buyers' focus and the impact of commodity prices on the evaluation.It provide valuable information in order to develop marketing strategies for high-quality sellers,improve the site structure,increasing sales.
Keywords/Search Tags:Association Rule, Genetic Algorithm, Improved Genetic Algorithm, User Evaluation
PDF Full Text Request
Related items