Wind Power Generation Prediction on a Large Real Life Dataset

The intermittent nature of wind power generation is becoming more of a problem as the percentage of wind energy used in the grid is increasing. We propose a data driven approach using machine learning methods to predict daily wind power generation output. A novel aspect of this work is the verification of the algorithm on a massive dataset with more than 500,000 observations. KeywordsWind Power; Renewable energy; Machine Learning ; Logistic regression