Applied Biomedical And Machine Learning Methods To Analyze Diabetes Data
【摘要】：Recently,the number of patients of diabetes grow rapidly. But there is not a good therapeutic schedule to cure diabetes. It may have some rules in its data. But when we analyze it by ourselves,the work is very difficult to complete. And when we use machine learning to analyze data about diabetes,we can find some new rules in it. It may help people to cure diabetes in future. In this paper,we choose cluster which is a kind of machine learning to find something about diabetes. And when we use biomedical methods to analyze it,we find that individual species of a lipid class may play significant roles in pathogenesis of type 2 diabetes. But when we use machine learning to analyze its gene expression,we can classify it as diabetic,pre-diabetic and normal.