Dissertation
Dissertation > Industrial Technology > Automation technology,computer technology > Automated basic theory > Artificial intelligence theory

Study on the Decision Tree Classification Algorithm and Its Application Based on Rough Set Theory

Author ZhouGuoJun
Tutor QinLiangZuo
School Guangxi University
Course Computer technology
Keywords Rough Set Theory Data Mining Data preprocessing Decision Tree Classification
CLC TP18
Type Master's thesis
Year 2011
Downloads 39
Quotes 0
Download Dissertation

Given the current lot of decision tree classification algorithm can not effectively deal with continuous attributes and missing values ??, based on rough set theory , decision tree classification algorithm , the purpose of the study is to construct a decision tree has a higher classification accuracy , so that the decision tree classifier The algorithm can effectively deal with continuous attributes and missing values ??. In order to achieve the purpose of the study , based on rough set theory in data preprocessing algorithm based on rough set theory , decision tree construction algorithm to study : less for incomplete decision table discretization algorithm proposed based on attributes important sexual discretization improved algorithm ; some flaws for ROUSTIDA algorithm , proposed an improved algorithm ROUSTIDA ; proposes a tree construction algorithm based on rough set theory , the algorithm attribute importance and approximate classification measure accuracy as constructing a decision tree attribute selection . On the basis of the algorithm and algorithm used in this paper , the design based on rough set theory , decision tree classification algorithm , the classification algorithm consists of three main steps : read the sample set and sample set pretreatment based rough set theory decision tree construction algorithm to construct decision trees , using the the PEP method of decision tree pruning . The classification algorithm can effectively deal with continuous attributes and missing values ??, higher classification accuracy of the decision tree can be constructed , the classification algorithm has a better performance is verified by experiment . Design of an electronic school supplies sales classification system , the system 's main function is to be classified in accordance with the sales of e-learning products . The system application of the decision tree construction algorithm to construct decision trees , and using the the PEP method of decision tree pruning . Using SQL Server 2000 VC 6.0 to achieve the system , the test results show that the system can be classified in accordance with the sales of e-learning products , giving the sales decision makers of e-learning products provide a certain value analysis method and decision support .

Related Dissertations
More Dissertations