Decreasing Error Rate of k-NN for Data classification using Double weighted function

For prediction of future class label or Variable value, there are two techniques respectively – Classification and Regression [2,5]. Data classification is a process between query point and training dataset to categorize query point. Using pre-labeled training dataset and classification algorithm, class label is assigned to new query point. Pre-labeled training dataset is taken from historical database of real world [6]. There are many data classification techniques exist like Decision Tree, Neural Network and Support Vector Machine etc. Among all, k-NN algorithm is simple and effective supervised algorithm for large dataset [7]. k-NN algorithm use all existing training datasets every time to classify new query point[3]. So it required more space to store all training dataset but give effective result. Here, k-NN algorithm is explained in details below. Falguni N. Patel Assistant Professor, Information Technology Department, Sardar Vallabhbhai Patel Institute of Technology, Vasad, DistAnand, Gujarat. falgunimanish19 @gmail.com