Research Journal of Pharmacy and Technology

K. Naresh ¹, N. Prabakaran ¹, R. Kannadasan ¹, P. Boominathan ¹

Affiliations
1 School of Computer Science and Engineering, VIT University, Vellore-632014, IN

Abstract

Data mining is the process of analyzing data from different perspectives and summarizing it into a useful information. In this paper we propose a different classification algorithm to identify the accuracy on diabetic data sets. The diabetic person has risk and leads to other disease such as blood vessel damage, blindness, heart diseases, nerve damage and kidney diseases. Diabetics also classified as two types such as type insulin diabetes and non-insulin dependent, diabetes is a disease in which the blood glucose increases which is due to the defects of secretion of insulin, or its action or both. Diabetes is a prolonged medical disease. In diabetes the cells of person produce insufficient amount of insulin or defective insulin or may insulin or may unable use insulin properly and efficiently that further leads to hyperglycemia and type-2 diabetes. We are proposing an efficient two level for classifying data. During initial phase we use training data for analyzing the optimality of dataset then new dataset is formed as optimal training dataset now we apply our classification mechanism on new diabetic datasets. The data mining methods and techniques will be explored to identify suitable methods and techniques for efficient classification on diabetic data set and in mining it in useful patterns.

Keywords

Data Mining, Diabetic Dataset, Classification, Naive Bayes Classification, Random Forest.

Full Text