机器学习算法在糖尿病预测中的应用  

Research on Diabetes Prediction Model Based on Machine Learning Algorithm

作　　者：贺其 赵岗 菊云霞 周薏岚 李敏 董琪 赵凯[3] HE Qi;ZHAO Gang;JU Yunxia;ZHOU Yilan;LI Min;DONG Qi;ZHAO Kai(School of Mathematics, Qilu Normal University, Jinan 250200, China;School of Information Science and Engineering,Shandong Normal University, Jinan 250014, China;School of Electrical Engineering and Automation, Qilu University of Technology(Shandong Academy of Sciences), Jinan 250353, China)

机构地区：[1]齐鲁师范学院数学学院,山东济南250200 [2]山东师范大学信息科学与工程学院,山东济南250014 [3]齐鲁工业大学(山东省科学院)电气工程与自动化学院,山东济南250353

出　　处：《贵州大学学报（自然科学版）》2019年第2期65-68,共4页Journal of Guizhou University:Natural Sciences

基　　金：山东省重点研发计划项目资助(2015GGH309003)

摘　　要：在很多领域利用机器学习的方法对数据进行分析、预测、判断具有非常重要的现实意义。将机器学习的算法应用到医学领域成为了研究的热点之一。糖尿病是多发病症,对是否患有糖尿病做出有效预测,意义重大。论文采用机器学习算法预测糖尿病,利用微软的Azure machine learning作为实验平台。采用了神经网络、逻辑回归、决策树、贝叶斯、支持向量机五种机器学习算法进行了预测,预测正确率分别是0.854,0.787,0.952,0.779,0.781。结果显示决策树预测效果最佳。在决策树预测的基础上对预测方法做出改进后,实验结果表明正确率提高了0.002。Using machine learning method to analyze, predict and judge in some fields, It is of great significance. The application of machine learning in medical field has become a hot topic of research. Diabetes is a common disease. It is of great significance to make effective prediction of diabetes. Machine learning algorithm was used to predict diabetes. Using Microsoft Azure machine learning as the experimental platform, a data set with 15000 samples was chosen. Each sample has 11 feature points, and 70% samples were used as training set and 30% as a test set. Neural network, logical regression, decision tree, Bayes, support vector machine were used to predict and the accuracy of prediction is 0.854, 0.787, 0.952, 0.779, 0.781 respectively. The prediction results show that the decision tree prediction is better. After further improvement of the prediction method, the experimental results show that the accuracy rate is increased by 0.002.

关 键 词：机器学习 糖尿病 决策树 AZURE MACHINE LEARNING 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] R587[自动化与计算机技术—控制科学与工程]