Adaptive locally connected recurrent unit (ALCRU)

Basit Öğe Kaydı
dc.authorid0000-0003-3903-7356
dc.authorid0000-0002-8649-6013
dc.contributor.authorÖzçelik, Şuayb Talhaen_US
dc.contributor.authorTek, Faik Borayen_US
dc.date.accessioned2025-09-15T07:00:04Z
dc.date.available2025-09-15T07:00:04Z
dc.date.issued2025-07-03
dc.departmentIşık Üniversitesi, Lisansüstü Eğitim Enstitüsü, Bilgisayar Mühendisliği Doktora Programıen_US
dc.departmentIşık University, School of Graduate Studies, Ph.D. in Computer Engineeringen_US
dc.description.abstractResearch has shown that adaptive locally connected neurons outperform their fully connected (dense) counterparts, motivating this study on the development of the Adaptive Locally Connected Recurrent Unit (ALCRU). ALCRU modifies the Simple Recurrent Neuron Model (SimpleRNN) by incorporating spatial coordinate spaces for input and hidden state vectors, facilitating the learning of parametric local receptive fields. These modifications add four trainable parameters per neuron, resulting in a minor increase in computational complexity. ALCRU is implemented using standard frameworks and trained with back-propagation-based optimizers. We evaluate the performance of ALCRU using diverse benchmark datasets, including IMDb for sentiment analysis, AdditionRNN for sequence modelling, and the Weather dataset for time-series forecasting. Results show that ALCRU achieves accuracy and loss metrics comparable to GRU and LSTM while consistently outperforming SimpleRNN. In particular, experiments with longer sequence lengths on AdditionRNN and increased input dimensions on IMDb highlight ALCRU’s superior scalability and efficiency in processing complex data sequences. In terms of computational efficiency, ALCRU demonstrates a considerable speed advantage over gated models like LSTM and GRU, though it is slower than SimpleRNN. These findings suggest that adaptive local connectivity enhances both the accuracy and efficiency of recurrent neural networks, offering a promising alternative to standard architectures.en_US
dc.description.versionPublisher's Versionen_US
dc.identifier.citationÖzçelik, Ş. T. & Tek, F. B. (2025). Adaptive locally connected recurrent unit (ALCRU). International Journal of Machine Learning and Cybernetics, 16(9), 6903-6914. doi:https://doi.org/10.1007/s13042-025-02652-7en_US
dc.identifier.doi10.1007/s13042-025-02652-7
dc.identifier.endpage6914
dc.identifier.issn1868-8071
dc.identifier.issn1868-808X
dc.identifier.issue9
dc.identifier.scopus2-s2.0-105009623173
dc.identifier.scopusqualityQ1
dc.identifier.startpage6903
dc.identifier.urihttps://hdl.handle.net/11729/6700
dc.identifier.urihttps://doi.org/10.1007/s13042-025-02652-7
dc.identifier.volume16
dc.identifier.wosWOS:001522431100001
dc.identifier.wosqualityQ3
dc.indekslendigikaynakScopusen_US
dc.indekslendigikaynakWeb of Scienceen_US
dc.indekslendigikaynakScience Citation Index Expanded (SCI-EXPANDED)en_US
dc.institutionauthorÖzçelik, Şuayb Talhaen_US
dc.institutionauthorid0000-0003-3903-7356
dc.language.isoenen_US
dc.peerreviewedYesen_US
dc.publicationstatusPublisheden_US
dc.publisherSpringer Science and Business Media Deutschland GmbHen_US
dc.relation.ispartofInternational Journal of Machine Learning and Cyberneticsen_US
dc.relation.publicationcategoryMakale - Uluslararası Hakemli Dergi - Öğrencien_US
dc.rightsinfo:eu-repo/semantics/openAccessen_US
dc.subjectAdaptiveen_US
dc.subjectFocusing neuronen_US
dc.subjectReceptive fielden_US
dc.subjectRNNen_US
dc.subjectBackpropagationen_US
dc.subjectBenchmarkingen_US
dc.subjectComplex networksen_US
dc.subjectData accuracyen_US
dc.subjectData handlingen_US
dc.subjectLong short-term memoryen_US
dc.subjectNetwork architectureen_US
dc.subjectNeuronsen_US
dc.subjectSentiment analysisen_US
dc.subjectTime series analysisen_US
dc.subjectCoordinate spaceen_US
dc.subjectHidden stateen_US
dc.subjectInput stateen_US
dc.subjectNeuron modelingen_US
dc.subjectReceptive fieldsen_US
dc.subjectSimple++en_US
dc.subjectSpatial coordinatesen_US
dc.subjectComputational efficiencyen_US
dc.subjectNeural-networksen_US
dc.subjectModelen_US
dc.subjectLSTMen_US
dc.titleAdaptive locally connected recurrent unit (ALCRU)en_US
dc.typeArticleen_US
dspace.entity.typePublicationen_US

Dosyalar

Orijinal paket
Listeleniyor 1 - 1 / 1
Küçük Resim Yok
İsim:
Adaptive_locally_connected_recurrent_unit_ALCRU.pdf
Boyut:
2.56 MB
Biçim:
Adobe Portable Document Format
İndir
Lisans paketi
Listeleniyor 1 - 1 / 1
Küçük Resim Yok
İsim:
license.txt
Boyut:
1.17 KB
Biçim:
Item-specific license agreed upon to submission
Açıklama:
İndir

Koleksiyon

Öğrenci Yayınları Makale Koleksiyonu
Lisansüstü Eğitim Enstitüsü Diğer Yayınlar Koleksiyonu
Scopus İndeksli Yayınlar Koleksiyonu
WoS İndeksli Yayınlar Koleksiyonu