Application of SVM as Classifier in Estimating Market Power Under Deregulated Electricity Market

This paper presents an approach to the application of Support Vector Machines (SVM) in estimating market power with the perspective of generation companies (Gencos). By choosing proper attributes and criterion for classification, SVM can be effectively used as a tool for market power estimation in the deregulated electricity market. In this paper, an analysis has been made to infer how effective the SVM technique is in predicting whether market power can be exercised by an entity or not. The person who is handling the technique i.e. Gencos decides the attributes chosen and the classifying criterion. Nodal Must Run Share Index is used in estimating market power. A simple three bus system with two generators and one load/two loads is chosen for the study. Performance of both linear and non-linear kernels is also compared.