AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain

<p dir="ltr">This article proposes a new deep learning framework for time series classification in the discrete cosine transform (DCT) domain with spectral enhancement and self-attention mechanisms. The time series signal is first partitioned into discrete segments. Each segment is r...

وصف كامل

محفوظ في:
التفاصيل البيبلوغرافية
المؤلف الرئيسي: Amine Haboub (22392097) (author)
مؤلفون آخرون: Hamza Baali (14603380) (author), Abdesselam Bouzerdoum (17900021) (author)
منشور في: 2025
الموضوعات:
الوسوم: إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
_version_ 1864513537623195648
author Amine Haboub (22392097)
author2 Hamza Baali (14603380)
Abdesselam Bouzerdoum (17900021)
author2_role author
author
author_facet Amine Haboub (22392097)
Hamza Baali (14603380)
Abdesselam Bouzerdoum (17900021)
author_role author
dc.creator.none.fl_str_mv Amine Haboub (22392097)
Hamza Baali (14603380)
Abdesselam Bouzerdoum (17900021)
dc.date.none.fl_str_mv 2025-06-01T00:00:00Z
dc.identifier.none.fl_str_mv 10.1109/tai.2025.3534141
dc.relation.none.fl_str_mv https://figshare.com/articles/journal_contribution/AttDCT_Attention-Based_Deep_Learning_Approach_for_Time_Series_Classification_in_the_DCT_Domain/30306052
dc.rights.none.fl_str_mv CC BY 4.0
info:eu-repo/semantics/openAccess
dc.subject.none.fl_str_mv Information and computing sciences
Artificial intelligence
Machine learning
Mathematical sciences
Applied mathematics
Attention
convolutional neural networks (CNN)
deep learning
discrete cosine transform (DCT)
time series classification (TSC)
dc.title.none.fl_str_mv AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
dc.type.none.fl_str_mv Text
Journal contribution
info:eu-repo/semantics/publishedVersion
text
contribution to journal
description <p dir="ltr">This article proposes a new deep learning framework for time series classification in the discrete cosine transform (DCT) domain with spectral enhancement and self-attention mechanisms. The time series signal is first partitioned into discrete segments. Each segment is rearranged into a matrix using a sliding window. The signal matrix is then transformed to spectral coefficients using a two-dimensional (2-D) DCT. This is followed by logarithmic contrast enhancement and spectral normalization to enhance the DCT coefficients. The resulting enhanced coefficient matrix serves as input to a deep neural network architecture comprising a self-attention layer, a multilayer convolutional neural network (CNN), and a fully connected multilayer perceptron (MLP) for classification. The AttDCT CNN model is evaluated and benchmarked on 13 different time series classification problems. The experimental results show that the proposed model outperforms state-of-the-art deep learning methods by an average of 2.1% in classification accuracy. It achieves higher classification accuracy on ten of the problems and similar results on the remaining three.</p><h2>Other Information</h2><p dir="ltr">Published in: IEEE Transactions on Artificial Intelligence<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/tai.2025.3534141" target="_blank">https://dx.doi.org/10.1109/tai.2025.3534141</a></p>
eu_rights_str_mv openAccess
id Manara2_715c8399e1a9aba0b6549c812329f46b
identifier_str_mv 10.1109/tai.2025.3534141
network_acronym_str Manara2
network_name_str Manara2
oai_identifier_str oai:figshare.com:article/30306052
publishDate 2025
repository.mail.fl_str_mv
repository.name.fl_str_mv
repository_id_str
rights_invalid_str_mv CC BY 4.0
spelling AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT DomainAmine Haboub (22392097)Hamza Baali (14603380)Abdesselam Bouzerdoum (17900021)Information and computing sciencesArtificial intelligenceMachine learningMathematical sciencesApplied mathematicsAttentionconvolutional neural networks (CNN)deep learningdiscrete cosine transform (DCT)time series classification (TSC)<p dir="ltr">This article proposes a new deep learning framework for time series classification in the discrete cosine transform (DCT) domain with spectral enhancement and self-attention mechanisms. The time series signal is first partitioned into discrete segments. Each segment is rearranged into a matrix using a sliding window. The signal matrix is then transformed to spectral coefficients using a two-dimensional (2-D) DCT. This is followed by logarithmic contrast enhancement and spectral normalization to enhance the DCT coefficients. The resulting enhanced coefficient matrix serves as input to a deep neural network architecture comprising a self-attention layer, a multilayer convolutional neural network (CNN), and a fully connected multilayer perceptron (MLP) for classification. The AttDCT CNN model is evaluated and benchmarked on 13 different time series classification problems. The experimental results show that the proposed model outperforms state-of-the-art deep learning methods by an average of 2.1% in classification accuracy. It achieves higher classification accuracy on ten of the problems and similar results on the remaining three.</p><h2>Other Information</h2><p dir="ltr">Published in: IEEE Transactions on Artificial Intelligence<br>License: <a href="https://creativecommons.org/licenses/by/4.0/deed.en" target="_blank">https://creativecommons.org/licenses/by/4.0/</a><br>See article on publisher's website: <a href="https://dx.doi.org/10.1109/tai.2025.3534141" target="_blank">https://dx.doi.org/10.1109/tai.2025.3534141</a></p>2025-06-01T00:00:00ZTextJournal contributioninfo:eu-repo/semantics/publishedVersiontextcontribution to journal10.1109/tai.2025.3534141https://figshare.com/articles/journal_contribution/AttDCT_Attention-Based_Deep_Learning_Approach_for_Time_Series_Classification_in_the_DCT_Domain/30306052CC BY 4.0info:eu-repo/semantics/openAccessoai:figshare.com:article/303060522025-06-01T00:00:00Z
spellingShingle AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
Amine Haboub (22392097)
Information and computing sciences
Artificial intelligence
Machine learning
Mathematical sciences
Applied mathematics
Attention
convolutional neural networks (CNN)
deep learning
discrete cosine transform (DCT)
time series classification (TSC)
status_str publishedVersion
title AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
title_full AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
title_fullStr AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
title_full_unstemmed AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
title_short AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
title_sort AttDCT: Attention-Based Deep Learning Approach for Time Series Classification in the DCT Domain
topic Information and computing sciences
Artificial intelligence
Machine learning
Mathematical sciences
Applied mathematics
Attention
convolutional neural networks (CNN)
deep learning
discrete cosine transform (DCT)
time series classification (TSC)