Paper Title
Hybridization of Svm-Knn Based Model Design for Short Term Load Forecast in Distribution System

Abstract
Load forecasting plays a vital role to enhance the efficiency of the distributed functions in distributed system. The paper represents the various aspects related to load forecasting system such as need of forecasting, issues to the load forecasting system. Earlier, threshold method was used for load balancing which suffers from many problems such as inconsistency in dataset led to the inefficient results. In order to overcome this issue the new technique for efficient load forecasting is also represented in this work which is based on KNN and SVM. The load forecasting dataset is firstly classified in various sections and then SVM is applied to train and test the dataset. The proposed work checks for patterns of data if any mismatch is found in the pattern then that entry is filled up on the basis of the results of the nearby entries. For the comparison purpose the normal SVM based load forecasting system is also analyzed and then compared with the results of the proposed work. The accuracy of the distributed system is evaluated to enhance by 90-95% as compared to the traditional single SVM based load forecasting technique. Keywords- Load Forecasting, K-Nearest Neighbor, Support Vector Machine, Threshold