Application of classification based data mining technique in diabetes care

The present research work relates data mining to medical informatics. The proposed work shows various models for each type of diabetic intervention and analysis is carried out using classification based data mining technique. The Area Under Curve (AUC) of ROC (Receiving Operating Characteristics) plots are calculated, the confusion matrix is formed, through which accuracy and cost of interventions have been evaluated. The AUC of ROC for all six modes of diabetic interventions are obtained and have been distinguished which mode of intervention is more appropriate. The accuracy and AUC of the model depend on the cost of model which is always inversely proportional to the cost of the model. Present analysis predicts that smoking cessation is the best intervention followed by exercise, diet, weight and drug for the diabetic control. Therefore, the results are quite impressive in predicting the diabetic intervention control resulting in high AUC of ROC and high accuracy with lowest cost.