Clustering of electricity price: an application to the Italian electricity market

Abstract

Analyzing the electricity price plays a vital role in market players in deregulated electricity markets. In this regard, proper clustering methods are more beneficial. In this paper, a comparative study of three major clustering algorithms, including K-Means, Fuzzy C-Means, and Hierarchical algorithm on the Italian National Single Price, has been carried out. Moreover, the impact of various parameters of the algorithms on the clustering results has been analyzed. The performance of the clustering methods has been compared through several clustering validity indexes, including the Silhouette index, Calinski-Harabasz index, and Davies-Bouldin indicator. Two distinct patterns consisting of working days and weekends (or holidays) are observable in the particular dataset.