Performance evaluation of activation functions in deep residual networks for short-term load forecasting

Short-Term Load Forecasting (STLF) is essential for ensuring efficient and reliable power system operations, requiring accurate predictions of electricity demand. Deep Residual Networks (DRNs), with their ability to mitigate gradient vanishing and model complex nonlinear relationships in load data,...

Full description

Bibliographic Details
Main Authors: Liu, Junchen, Ahmad, Faisul Arif, Samsudin, Khairulmizam, Hashim, Fazirulhisyam, Ab Kadir, Mohd Zainal Abidin
Format: Article
Language:English
Published: Institute of Electrical and Electronics Engineers 2025
Online Access:http://psasir.upm.edu.my/id/eprint/118663/
http://psasir.upm.edu.my/id/eprint/118663/1/118663.pdf
_version_ 1848867567737765888
author Liu, Junchen
Ahmad, Faisul Arif
Samsudin, Khairulmizam
Hashim, Fazirulhisyam
Ab Kadir, Mohd Zainal Abidin
author_facet Liu, Junchen
Ahmad, Faisul Arif
Samsudin, Khairulmizam
Hashim, Fazirulhisyam
Ab Kadir, Mohd Zainal Abidin
author_sort Liu, Junchen
building UPM Institutional Repository
collection Online Access
description Short-Term Load Forecasting (STLF) is essential for ensuring efficient and reliable power system operations, requiring accurate predictions of electricity demand. Deep Residual Networks (DRNs), with their ability to mitigate gradient vanishing and model complex nonlinear relationships in load data, have emerged as a powerful tool for STLF. This study evaluates the performance of various activation functions within DRN models, focusing on their impact on predictive precision and generalization. Experiments were conducted using the DRN architecture for STLF on two distinct datasets: ISO-NE and Malaysia. The findings demonstrate that activation functions significantly influence the predictive performance of DRN-based STLF models. Specifically, the DRN model using Swish achieved the best results on the ISO-NE dataset (Mean Absolute Percentage Error, MAPE = 1.3806%), while the DRN model with Hyperbolic Tangent (Tanh) excelled on the Malaysia dataset (MAPE = 4.9809%). These results underscore the importance of aligning activation function selection with dataset characteristics to optimize the performance of DRN models in STLF. This study provides valuable insights for advancing STLF research and guiding practical applications in load forecasting.
first_indexed 2025-11-15T14:38:33Z
format Article
id upm-118663
institution Universiti Putra Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T14:38:33Z
publishDate 2025
publisher Institute of Electrical and Electronics Engineers
recordtype eprints
repository_type Digital Repository
spelling upm-1186632025-07-22T02:47:05Z http://psasir.upm.edu.my/id/eprint/118663/ Performance evaluation of activation functions in deep residual networks for short-term load forecasting Liu, Junchen Ahmad, Faisul Arif Samsudin, Khairulmizam Hashim, Fazirulhisyam Ab Kadir, Mohd Zainal Abidin Short-Term Load Forecasting (STLF) is essential for ensuring efficient and reliable power system operations, requiring accurate predictions of electricity demand. Deep Residual Networks (DRNs), with their ability to mitigate gradient vanishing and model complex nonlinear relationships in load data, have emerged as a powerful tool for STLF. This study evaluates the performance of various activation functions within DRN models, focusing on their impact on predictive precision and generalization. Experiments were conducted using the DRN architecture for STLF on two distinct datasets: ISO-NE and Malaysia. The findings demonstrate that activation functions significantly influence the predictive performance of DRN-based STLF models. Specifically, the DRN model using Swish achieved the best results on the ISO-NE dataset (Mean Absolute Percentage Error, MAPE = 1.3806%), while the DRN model with Hyperbolic Tangent (Tanh) excelled on the Malaysia dataset (MAPE = 4.9809%). These results underscore the importance of aligning activation function selection with dataset characteristics to optimize the performance of DRN models in STLF. This study provides valuable insights for advancing STLF research and guiding practical applications in load forecasting. Institute of Electrical and Electronics Engineers 2025 Article PeerReviewed text en cc_by_4 http://psasir.upm.edu.my/id/eprint/118663/1/118663.pdf Liu, Junchen and Ahmad, Faisul Arif and Samsudin, Khairulmizam and Hashim, Fazirulhisyam and Ab Kadir, Mohd Zainal Abidin (2025) Performance evaluation of activation functions in deep residual networks for short-term load forecasting. IEEE Access, 13. art. no. undefined. pp. 78618-78633. ISSN 2169-3536 https://ieeexplore.ieee.org/document/10980244/ 10.1109/ACCESS.2025.3565798
spellingShingle Liu, Junchen
Ahmad, Faisul Arif
Samsudin, Khairulmizam
Hashim, Fazirulhisyam
Ab Kadir, Mohd Zainal Abidin
Performance evaluation of activation functions in deep residual networks for short-term load forecasting
title Performance evaluation of activation functions in deep residual networks for short-term load forecasting
title_full Performance evaluation of activation functions in deep residual networks for short-term load forecasting
title_fullStr Performance evaluation of activation functions in deep residual networks for short-term load forecasting
title_full_unstemmed Performance evaluation of activation functions in deep residual networks for short-term load forecasting
title_short Performance evaluation of activation functions in deep residual networks for short-term load forecasting
title_sort performance evaluation of activation functions in deep residual networks for short-term load forecasting
url http://psasir.upm.edu.my/id/eprint/118663/
http://psasir.upm.edu.my/id/eprint/118663/
http://psasir.upm.edu.my/id/eprint/118663/
http://psasir.upm.edu.my/id/eprint/118663/1/118663.pdf