Justifying Short-Term Load Forecasts Obtained with the Use of Neural Models

There is a lot of research on the neural models used for short-term load forecasting (STLF), which crucial improving sustainable operation energy systems with increasing technical, economic, and environmental requirements. Neural networks are computationally powerful; however, lack clear, readable trustworthy justification STLF obtained using such serious problem that needs to be tackled. The article proposes an approach based local interpretable model-agnostic explanations (LIME) method supports reliable premises justifying explaining forecasts. use proposed makes it possible improve reliability heuristic experimental modeling processes, results difficult interpret. Explaining may facilitate selection improvement STLF, while contributing better understanding broadening knowledge experience supporting enhancement security forecasts simplifying dispatch decisions.