Filtreler

2023 - 20252
Filtreleri Sıfırla

Ayarlar

Yazar: Özçelik, Şuayb Talha × Bölüm: Işık Üniversitesi, Lisansüstü Eğitim Enstitüsü, Bilgisayar Mühendisliği Doktora Programı ×

Arama Sonuçları

Listeleniyor 1 - 2 / 2
  • Küçük Resim
    Yayın
    Adaptive locally connected recurrent unit (ALCRU)
    (Springer Science and Business Media Deutschland GmbH, 2025-07-03) Özçelik, Şuayb Talha; Tek, Faik Boray
    Research 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.
  • Küçük Resim
    Yayın
    Transforming tourism experience: AI-based smart travel platform
    (Association for Computing Machinery, 2023) Yöndem, Meltem Turhan; Özçelik, Şuayb Talha; Caetano, Inés; Figueiredo, José; Alves, Patrícia; Marreiros, Goreti; Bahtiyar, Hüseyin; Yüksel, Eda; Perales, Fernando
    In this paper, we propose the development of a novel personalized tourism platform incorporating artificial intelligence (AI) and augmented reality (AR) technologies to enhance the smart tourism experience. The platform utilizes various data sources, including travel history, user activity, and personality assessments, combined with machine learning algorithms to generate tailored travel recommendations for individual users. We implemented fundamental requirements for the platform: secure user identification using blockchain technology and provision of personalized services based on user interests and preferences. By addressing these requirements, the platform aims to increase tourist satisfaction and improve the efficiency of the tourism industry. In collaboration with various universities and companies, this multinational project aims to create a versatile platform that can seamlessly integrate new smart tourism units, providing an engaging, educational, and enjoyable experience for users.