BP算法与C4.5算法在乳腺癌诊断中的比较分析  

Comparison and analysis of BP algorithm and C4.5 algorithm in diagnosis of breast cancer

作　　者：杨云[1] 董雪[1] 齐勇[1] 

机构地区：[1]陕西科技大学电气与信息工程学院

出　　处：《陕西科技大学学报（自然科学版）》2015年第3期163-166,172,共5页Journal of Shaanxi University of Science & Technology

基　　金：陕西省科技厅科学技术研究发展计划项目(2014K15-03-06);西安市科技计划项目(NC1403(2);NC1319(1))

摘　　要：目前数据挖掘技术被大量应用于医学领域,进行疾病诊断。针对乳腺癌发病率不断升高,为辅助医生做出诊断决策,采用具有优秀学习能力的人工神经网络中的BP算法与决策树中的C4.5算法来分析乳腺癌数据,对乳腺癌肿瘤类型进行诊断预测,并对这两种算法建立的分类器性能进行比较分析,研究发现BP算法与C4.5算法都能对乳腺癌类型作出诊断预测,但在分类器的评估中发现BP分类器的性能优于C4.5分类器.At present, a large number of data mining techniques are applied in the medical field for diagnosis of the disease. In view of breast cancer incidence rate increased continuously, in order to assist the doctor make a diagnosis decision-making, the BP neural network algorithm and C4.5 decision tree algorithm which have excellent learning ability were used to analyze the data of breast cancer. In order to predict the type of breast cancer tumor, and the performance of the two established classifiers were compared. The study found that the performance of BP classifier in the diagnosis of breast cancer is better than C4.5 classifier.

关 键 词：乳腺癌分类 BP算法 C4.5算法 分类器性能 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]