Skip to content

AlphaGrad: Normalized Gradient Descent for Adaptive Multi-loss Functions in EEG-based Motor Imagery Classification (IEEE Journal of Biomedical and Health Informatics)

Notifications You must be signed in to change notification settings

xydxdy/AlphaGrad

Repository files navigation

AlphaGrad: Normalized Gradient Descent for Adaptive Multi-loss Functions in EEG-based Motor Imagery Classification

DOI

PyTorch Supplementary

AlphaGrad

In this study, we propose AlphaGrad, a novel adaptive loss blending strategy for optimizing multi-task learning (MTL) models in motor imagery (MI)-based electroencephalography (EEG) classification. AlphaGrad is the first method to automatically adjust multi-loss functions with differing metric scales, including mean square error, cross-entropy, and deep metric learning, within the context of MI-EEG. We evaluate AlphaGrad using two state-of-the-art MTL-based neural networks, MIN2Net and FBMSNet, across four benchmark datasets. Experimental results show that AlphaGrad consistently outperforms existing strategies such as AdaMT, GradApprox, and fixed-weight baselines in classification accuracy and training stability. Compared to baseline static weighting, AlphaGrad achieves over 10% accuracy improvement on subject-independent MI tasks when evaluated on the largest benchmark dataset. Furthermore, AlphaGrad demonstrates robust adaptability across various EEG paradigms—including steady-state visually evoked potential (SSVEP) and event-related potential (ERP), making it broadly applicable to brain-computer interface (BCI) systems. We also provide gradient trajectory visualizations highlighting AlphaGrad’s ability to maintain training stability and avoid local minima. These findings underscore AlphaGrad’s promise as a general-purpose solution for adaptive multi-loss optimization in biomedical time-series learning.

Trajectory of gradient descent (GD) for a multi-task objective

Multi-task Objective

Citation

To cited our paper

R. Chaisaen, P. Autthasan, A. Ditthapron and T. Wilaiprasitporn, "AlphaGrad: Normalized Gradient Descent for Adaptive Multi-loss Functions in EEG-based Motor Imagery Classification," in IEEE Journal of Biomedical and Health Informatics, doi: 10.1109/JBHI.2025.3572197.

@ARTICLE{11008918,
  author={Chaisaen, Rattanaphon and Autthasan, Phairot and Ditthapron, Apiwat and Wilaiprasitporn, Theerawit},
  journal={IEEE Journal of Biomedical and Health Informatics}, 
  title={AlphaGrad: Normalized Gradient Descent for Adaptive Multi-loss Functions in EEG-based Motor Imagery Classification}, 
  year={2025},
  volume={},
  number={},
  pages={1-13},
  keywords={Electroencephalography;Training;Feature extraction;Measurement;Multitasking;Brain modeling;Adaptation models;Motors;Filter banks;Classification algorithms;Brain-computer interfaces;multi-task learning;adaptive loss blending;motor imagery EEG;normalized gradient descent},
  doi={10.1109/JBHI.2025.3572197}}

About

AlphaGrad: Normalized Gradient Descent for Adaptive Multi-loss Functions in EEG-based Motor Imagery Classification (IEEE Journal of Biomedical and Health Informatics)

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages