Robust Probabilistic Load Forecasting for a Single Household: A Comparative Study from SARIMA to Transformers on the REFIT Dataset

TL;DR

This study compares various probabilistic load forecasting models for households using the REFIT dataset, highlighting the superiority of the Temporal Fusion Transformer in accuracy and uncertainty quantification.

Contribution

It introduces a comprehensive evaluation of classical, machine learning, and deep learning models for household load forecasting, emphasizing the effectiveness of the TFT architecture.

Findings

TFT achieved the best point forecast accuracy (RMSE 481.94).

LSTM provided well-calibrated probabilistic forecasts.

Classical models like SARIMA and Prophet failed to capture non-linear behaviors.

Abstract

Probabilistic forecasting is essential for modern risk management, allowing decision-makers to quantify uncertainty in critical systems. This paper tackles this challenge using the volatile REFIT household dataset, which is complicated by a large structural data gap. We first address this by conducting a rigorous comparative experiment to select a Seasonal Imputation method, demonstrating its superiority over linear interpolation in preserving the data's underlying distribution. We then systematically evaluate a hierarchy of models, progressing from classical baselines (SARIMA, Prophet) to machine learning (XGBoost) and advanced deep learning architectures (LSTM). Our findings reveal that classical models fail to capture the data's non-linear, regime-switching behavior. While the LSTM provided the most well-calibrated probabilistic forecast, the Temporal Fusion Transformer (TFT) emerged as the superior all-round model, achieving the best point forecast accuracy (RMSE 481.94) and producing safer, more cautious prediction intervals that effectively capture extreme volatility.