Dataset for An Application of Machine Learning for a Smart Grid Resource Allocation Problem

The main dataset from a year’s of running the SGRA on the HPC (Summit) environment is divided into two sub-datasets. The two sub-datasets are used to develop machine learning methods to obtain relationships between aggregator profits and customer loads, and electricity prices, respectively. The first and second sub-datasets consist of 5,555 and 365 observations, respectively. Each dataset is divided into two groups: training data (75%) and test data (25%).