VIEW THE RECORDING HERE.

A method to select the best load forecasting tool
Presenter: Yiyan Li, NC State Post Doc

Load forecasting at the distribution level is becoming more difficult and faces different forecasting requirements and data availability. As a result, no single model can handle all the forecasting problems. FREEDM researchers developed a forecasting model selection framework based on meta learning that can automatically identify the best forecasting model among candidates for each forecasting problem. In this way, every forecasting problem can find its optimal or near-optimal solution, so that the overall forecasting accuracy is improved.