Neural basis expansion analysis with exogenous variables: Forecasting electricity prices with NBEATSx

TL;DR

The paper introduces NBEATSx, an enhanced neural network model that incorporates exogenous variables for improved electricity price forecasting, achieving state-of-the-art accuracy and interpretability across diverse markets.

Contribution

NBEATSx extends the NBEATS model by integrating exogenous variables, enabling better forecasting accuracy and interpretability in electricity price prediction tasks.

Findings

Achieves nearly 20% improvement over original NBEATS

Outperforms established statistical and machine learning methods by up to 5%

Provides interpretable decomposition of time series components

Abstract

We extend the neural basis expansion analysis (NBEATS) to incorporate exogenous factors. The resulting method, called NBEATSx, improves on a well performing deep learning model, extending its capabilities by including exogenous variables and allowing it to integrate multiple sources of useful information. To showcase the utility of the NBEATSx model, we conduct a comprehensive study of its application to electricity price forecasting (EPF) tasks across a broad range of years and markets. We observe state-of-the-art performance, significantly improving the forecast accuracy by nearly 20% over the original NBEATS model, and by up to 5% over other well established statistical and machine learning methods specialized for these tasks. Additionally, the proposed neural network has an interpretable configuration that can structurally decompose time series, visualizing the relative impact of trend and seasonal components and revealing the modeled processes' interactions with exogenous factors. To assist related work we made the code available in https://github.com/cchallu/nbeatsx.