An Advance Tree Adaptive Data Classification for the Diabetes Disease Prediction

Data mining is one of the emerging area in the field of computer science it's enable to deal with large dataset with different characteristic. In the current scenario it is used in every field like Medical. Education, Agriculture etc., but in the past few decades use of data mining approaches is increasing exponentially because it required prediction based on data for quick decision. Sometimes it is very challenging to predict accurately on large study data. Classification and observing them is one of the proper solution which driven by algorithms. In this paper a proposed algorithm is given which take advantage of partitioning based on tree, further working with adaptive SVM approach for classification. The proposed architecture used pre-processing under sampling SMORT which enable in pruning the data. The approach is experimented using the Weka tool on diabetic dataset and compared with traditional tree based RF, RT and J48 Approach. The observed outcome shows the efficiency of proposed algorithm over the traditional solution of processing diabetic data and finding efficient classification from it.