Statistical Approach to Agglomerative Hierarchical Clustering on Large Datasets: A Case Study of Energy Community in Italy

As the interest in Energy Communities (EC) increases, the identification of user segments potentially interested in joining them becomes more important. However, the energy consumption profile of users is extremely variable and makes it difficult to formulate a precise analysis of the real profitability of joining the EC. Agglomerative clustering techniques can be useful for identifying an a priori unknown number of user segments, but the high number of necessary recordings prevents its use. A case study is presented, and a method based on stratified multiple sampling is shown for the application of Agglomerative Hierarchical Clustering to statistically significant samples extracted from large datasets, with a significant reduction in computational load.