基于WEKA数据挖掘平台的医学数据分类及肾病早期预测

Medical Data Classification and Early-prediction of Nephropathy Based on WEKA Platform

目的:随着医院信息化建设的发展,如何从海量医院临床信息挖掘出隐藏的关联数据,为患者提供早期辅助诊断,是目前智能医疗诊断领域热门的研究方向。方法:在WEKA数据挖掘平台的基础上,对慢性肾脏疾病数据进行挖掘分析,并比较不同挖掘算法的分类准确性。结果:对比实验结果表明,较于其他分类器,Random Forest分类器对慢性肾脏疾病数据集具有较好的分类准确性。结论:基于WEKA数据挖掘平台的医学数据最适合肾脏疾病诊断和预测的算法,为后期医疗行业的大数据分析及挖掘提供新思路。

Objective： With the development of hospital information construction, how to extract the hidden data from massive clinical information of hospitals so as to provide early diagnosis for patients is the popular research direction in the field of intelligent medical diagnosis. Methods： Based on the WEKA data mining platform, the data of chronic kidney disease is analyzed in paper and the classification accuracy of different mining algorithms are compared. Results： Compared with other classifiers model, Random Forest classifier has a good classification accuracy for chronic kidney disease. Conclusion： Medical data classification and early-prediction of nephropathy based on WEKA platform analyzes the most suitable algorithm for the diagnosis and prediction of kidney disease, and provides new ideas for the analysis and mining of big data in the later medical industry.