Study On Coronary Heart Disease Classification By Rough Set And Decision Tree Algorithm

Computer aided coronary heart disease diagnosis was a conjunction study point in medical and computer area. Coronary heart disease cases is a kind of data that contains huge amount of implicate information. The object of data mining is methods which searching implicated information from data, and establishes aided decision model. Based on the above background, data mining used for coronary heart disease analysis was meaningful.Classification is a main analysis mean in data mining. It can establish classification model and generates classification rules by analyzing data and can be used to analyze new data. Classification includes many kinds of methods: decision tree, association rule, bayes, neural network, genetic algorithm and so on.The analysis object of the paper was a group of coronary heart disease data. Rough set and decision tree (C4.5 algorithm) were selected as analysis methods. The study mainly focused on processing procedure, parameters design and results,compared the fitness of the two algorithms on coronary heart disease.The paper mainly worked on:1. Trimmed coronary heart disease data.2. Designed contents of experiment.3. Compared classification procedure and result.Through study, different classification rule set were acquired by rough set and C4.5 decision tree:1. The classification accuracy rate of C4.5 was higher than that of rough set.2. The number of rules gotten by the C4.5 decision tree algorithm was larger than that of rough set.3. C4.5 rules were much readable than rough set rules.