Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification
Deep learning methods have proved a promising performance for electroencephalography-based brain-computer interfaces (EEG-BCI). It is particularly encouraging that a subject-independent model can be trained using a large amount of other subjects’ data. Transfer learning methods such as adaptation or...
Saved in:
Main Authors: | , , |
---|---|
Other Authors: | |
Format: | Article |
Language: | English |
Published: |
2024
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/179921 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-179921 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-1799212024-09-03T01:32:50Z Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification R, Vishnupriya Robinson, Neethu M, Ramasubba Reddy School of Computer Science and Engineering Computer and Information Science Brain-computer interface Electroencephalography Deep learning methods have proved a promising performance for electroencephalography-based brain-computer interfaces (EEG-BCI). It is particularly encouraging that a subject-independent model can be trained using a large amount of other subjects’ data. Transfer learning methods such as adaptation or fine-tuning can be used on the pre-trained model to improve the performance. This study examined the influence of fine-tuning on the subject-independent model for EEG-based motor imagery (MI) classification using a genetic algorithm (GA). The proposed method is evaluated on the binary class MI dataset from the Korea University EEG dataset. Results show that the proposed GA-based fine-tuning approach statistically improved the average classification accuracy of the baseline model from 84.46% to 87.29%. More interestingly, our approach shows significant improvement in cases where the performance of the baseline model is poor after fine-tuning using other approaches. Further, layer-wise relevance propagation (LRP) is used to analyze the adapted models to gain a deeper understanding of the neurophysiological explanations underlying the model’s decision. 2024-09-03T01:32:49Z 2024-09-03T01:32:49Z 2024 Journal Article R, V., Robinson, N. & M, R. R. (2024). Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification. Brain-Computer Interfaces, 1-12. https://dx.doi.org/10.1080/2326263X.2024.2347790 2326-263X https://hdl.handle.net/10356/179921 10.1080/2326263X.2024.2347790 2-s2.0-85192143238 1 12 en Brain-Computer Interfaces © 2024 Informa UK Limited, trading as Taylor & Francis Group. All rights reserved. |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
Computer and Information Science Brain-computer interface Electroencephalography |
spellingShingle |
Computer and Information Science Brain-computer interface Electroencephalography R, Vishnupriya Robinson, Neethu M, Ramasubba Reddy Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
description |
Deep learning methods have proved a promising performance for electroencephalography-based brain-computer interfaces (EEG-BCI). It is particularly encouraging that a subject-independent model can be trained using a large amount of other subjects’ data. Transfer learning methods such as adaptation or fine-tuning can be used on the pre-trained model to improve the performance. This study examined the influence of fine-tuning on the subject-independent model for EEG-based motor imagery (MI) classification using a genetic algorithm (GA). The proposed method is evaluated on the binary class MI dataset from the Korea University EEG dataset. Results show that the proposed GA-based fine-tuning approach statistically improved the average classification accuracy of the baseline model from 84.46% to 87.29%. More interestingly, our approach shows significant improvement in cases where the performance of the baseline model is poor after fine-tuning using other approaches. Further, layer-wise relevance propagation (LRP) is used to analyze the adapted models to gain a deeper understanding of the neurophysiological explanations underlying the model’s decision. |
author2 |
School of Computer Science and Engineering |
author_facet |
School of Computer Science and Engineering R, Vishnupriya Robinson, Neethu M, Ramasubba Reddy |
format |
Article |
author |
R, Vishnupriya Robinson, Neethu M, Ramasubba Reddy |
author_sort |
R, Vishnupriya |
title |
Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
title_short |
Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
title_full |
Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
title_fullStr |
Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
title_full_unstemmed |
Genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
title_sort |
genetic algorithm based deep learning model adaptation for improvising the motor imagery classification |
publishDate |
2024 |
url |
https://hdl.handle.net/10356/179921 |
_version_ |
1814047353233997824 |