Evaluating the Effectiveness and Efficiency of Hyperparameter Optimization Algorithms for Short-Term Electricity Load Forecasting

Abstract

Accurate forecasting of electrical demand is essential for maintaining a stable and reliable power grid, op- optimizing the allocation of energy resources, and promoting efficient energy consumption practices. This study investigates the effectiveness of five hyperparameter optimization (HPO) algorithms - Random Search, Covariance Matrix Adaptation Evolution Strategy (CMA-ES), Bayesian Optimization, Particle Swarm Optimization (PSO), and Nevergrad Optimizer (NGOpt) across univariate and multivariate Short-Term Load Forecasting (STLF) tasks. Using the Panama Electricity dataset (n = 48,049), we evaluate the performance of HPO algorithms on a surrogate forecasting algorithm, XGBoost, in terms of accuracy (i.e., MAPE, R2) and runtime. Performance plots visualize these metrics across varying sample sizes from 1,000 to 20,000, and Kruskal-Wallis tests assess the statistical significance of the performance differences. Results reveal significant runtime advantages for HPO algorithms over Random Search. In univariate models, Bayesian optimization exhibited the lowest accuracy among the tested methods. This study offers valuable insights for optimizing XGBoost in the context of STLF and identifies areas for future research. © 2025 IEEE.