TABULAR TRANSFORMER ARCHITECTURE WITH OPTUNA OPTIMIZATION FOR EARLY DIAGNOSIS OF ALZHEIMER'S DISEASE

Timur Lale

doi:10.17780/ksujes.1772927

TR EN

ALZHEİMER HASTALIĞININ ERKEN TANISI İÇİN OPTUNA OPTİMİZASYONLU TABULAR TRANSFORMER MİMARİSİ

Öz

Alzheimer hastalığı (AH), dünya çapında demansın en yaygın nedeni olan ilerleyici bir nörodejeneratif hastalıktır. Bu çalışmada, Alzheimer hastalığının erken teşhisi için Optuna algoritması ile optimize edilmiş bir Tabular Transformer mimarisi önerilmiştir. Kaggle platformundan elde edilen veri seti, 2.149 hastaya ait demografik, klinik, laboratuvar ve nöropsikolojik değerlendirmeler dahil olmak üzere 33 farklı özellik içermektedir. Pearson korelasyon analizi, fonksiyonel kapasite ölçümlerinin, günlük yaşam aktivitelerinin ve MMSE puanlarının tanı ile güçlü negatif korelasyonlar gösterdiğini ortaya koymuştur. Model boyutu, dikkat başlıkları sayısı, kodlayıcı katman sayısı, ileri beslemeli ağ boyutu, bırakma oranı, öğrenme oranı, L2 düzenleme katsayısı ve parti boyutu gibi kritik hiperparametreler, ağaç yapılı Parzen Tahmincisi algoritması kullanılarak otomatik olarak optimize edilmiştir. Model %94,65 doğruluk, %92,72 hassasiyet, %92,11 duyarlılık, %92,41 F1 puanı ve %96,04 özgüllük elde etmiştir. Önerilen yöntem, klasik makine öğrenimi yöntemleriyle karşılaştırıldığında tüm performans metriklerinde üstün sonuçlar göstermiştir. Yüksek özgüllüğü, klinik uygulamada yanlış pozitif tanıları en aza indirerek gereksiz test maliyetlerini azaltma potansiyeli sunmaktadır.

Anahtar Kelimeler

TABULAR TRANSFORMER ARCHITECTURE WITH OPTUNA OPTIMIZATION FOR EARLY DIAGNOSIS OF ALZHEIMER'S DISEASE

Abstract

Alzheimer's disease (AD) represents the leading cause of dementia globally and is characterized by progressive neurodegeneration. In this study, a Tabular Transformer architecture optimized with the Optuna algorithm is proposed for the early diagnosis of Alzheimer's disease. The dataset obtained from the Kaggle platform contains 33 different features, including demographic, clinical, laboratory, and neuropsychological assessments from 2,149 patients. Pearson correlation analysis revealed that functional capacity measurements, activities of daily living, and MMSE scores exhibited strong negative correlations with diagnosis. Critical hyperparameters such as model size, number of attention heads, number of encoder layers, feedforward network size, dropout rate, learning rate, L2 regularization coefficient, and batch size were automatically optimized using the tree-structured Parzen Estimator algorithm. The model achieved 94.65% accuracy, 92.72% precision, 92.11% sensitivity, 92.41% F1-score, and 96.04% specificity. In comparison with classical machine learning methods, the proposed method demonstrated superior results in all performance metrics. Its high specificity offers the potential to reduce unnecessary testing costs by minimizing false-positive diagnoses in clinical practice.

Keywords

Ethical Statement

There is no need to obtain permission from the ethics committee for the article prepared There is no conflict of interest with any person / institution in the article prepared

References

AbdElminaam, D. S., Madbouly, M. M., Farag, M. S., Gomaa, I. A., Abd-Elghany Zeid, M., & Abualigah, L. (2023). ML_Alzheimer: Alzheimer Disease Prediction Using Machine Learning. 2023 International Mobile, Intelligent, and Ubiquitous Computing Conference (MIUCC), 409–414. IEEE. https://doi.org/10.1109/MIUCC58832.2023.10278361
Acharya, H., Mehta, R., & Kumar Singh, D. (2021). Alzheimer Disease Classification Using Transfer Learning. 2021 5th International Conference on Computing Methodologies and Communication (ICCMC), 1503–1508. IEEE. https://doi.org/10.1109/ICCMC51019.2021.9418294
Akiba, T., Sano, S., Yanase, T., Ohta, T., & Koyama, M. (2019). Optuna. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, 2623–2631. New York, NY, USA: ACM. https://doi.org/10.1145/3292500.3330701
Alp, S., Akan, T., Bhuiyan, Md. S., Disbrow, E. A., Conrad, S. A., Vanchiere, J. A., … Bhuiyan, M. A. N. (2024). Joint transformer architecture in brain 3D MRI classification: its application in Alzheimer’s disease classification. Scientific Reports, 14(1), 8996. https://doi.org/10.1038/s41598-024-59578-3
Asaduzzaman, M., Alom, Md. K., & Karim, Md. E. (2025). ALZENET: Deep learning-based early prediction of Alzheimer’s disease through magnetic resonance imaging analysis. Telematics and Informatics Reports, 17, 100189. https://doi.org/10.1016/j.teler.2025.100189
Aslan, E., & Özüpak, Y. (2025). Comparison of machine learning algorithms for automatic prediction of Alzheimer disease. Journal of the Chinese Medical Association, 88(2), 98–107. https://doi.org/10.1097/JCMA.0000000000001188
C R, N., M, K., & K, S. (2023). Classifying the stages of Alzheimer’s disease by using multi layer feed forward neural network. Procedia Computer Science, 218, 1845–1856. https://doi.org/10.1016/j.procs.2023.01.162
Chitradevi, D., Prabha, S., & Alex Daniel Prabhu. (2021). Diagnosis of Alzheimer disease in MR brain images using optimization techniques. Neural Computing and Applications, 33(1), 223–237. https://doi.org/10.1007/s00521-020-04984-7

Dai, Z., & Huang, T. (2025). A comparative study of ML based predictive models for Alzheimer disease prediction. Multiscale and Multidisciplinary Modeling, Experiments and Design, 8(8), 342. https://doi.org/10.1007/s41939-025-00931-y
Desai, M. B., Kumar, Y., & Pandey, S. (2024). Efficient Approach for Diagnosis and Detection of Alzheimer Diseases Using Deep Learning. 2024 International Conference on Advances in Computing Research on Science Engineering and Technology (ACROSET), 1–5. IEEE. https://doi.org/10.1109/ACROSET62108.2024.10743886
Dhanusha, C., & Senthil Kumar, A. V. (2021). Deep Recurrent Q Reinforcement Learning model to Predict the Alzheimer Disease using Smart Home Sensor Data. IOP Conference Series: Materials Science and Engineering, 1074(1), 012014. https://doi.org/10.1088/1757-899X/1074/1/012014
Dubey, Y., Bhongade, A., Palsodkar, P., & Fulzele, P. (2024). Efficient Explainable Models for Alzheimer’s Disease Classification with Feature Selection and Data Balancing Approach Using Ensemble Learning. Diagnostics, 14(24), 2770. https://doi.org/10.3390/diagnostics14242770
Ghaffar Nia, N., Kaplanoglu, E., & Nasab, A. (2023). Evaluation of artificial intelligence techniques in disease diagnosis and prediction. Discover Artificial Intelligence, 3(1), 5. https://doi.org/10.1007/s44163-023-00049-5
Hendrycks, D., & Gimpel, K. (2023). Gaussian Error Linear Units (GELUs).
Hu, Z., Wang, Z., Jin, Y., & Hou, W. (2023). VGG-TSwinformer: Transformer-based deep learning model for early Alzheimer’s disease prediction. Computer Methods and Programs in Biomedicine, 229, 107291. https://doi.org/10.1016/j.cmpb.2022.107291
Huang, X., Khetan, A., Cvitkovic, M., & Karnin, Z. (2020). TabTransformer: Tabular Data Modeling Using Contextual Embeddings.
Hussain, S., Shah, B., Khan, A., & Tanvir, S. (2025). A Novel Prediction Model for Alzheimer Classification Using Deep Learning. https://doi.org/10.1007/978-3-031-89813-6_15
Joon, D., Kumar, R., Gupta, M., & Obaid, A. J. (2024). A comprehensive Analysis on Diagnosis of Alzheimer Disease Using Generative Adversarial Network. 2024 International Conference on Advances in Computing, Communication and Applied Informatics (ACCAI), 1–6. IEEE. https://doi.org/10.1109/ACCAI61061.2024.10602112
Kamal, M. S., & Farhana Nimmy, S. (2024). Interpretable Transformers for Alzheimer Disease Diagnosis on Multi-modal Data. 2024 International Joint Conference on Neural Networks (IJCNN), 1–8. IEEE. https://doi.org/10.1109/IJCNN60899.2024.10651416
Kanna, R. K., Mutheeswaran, U., Ramya, V. S., Gomalavalli, R., Hema, L. K., & Ambikapathy, A. (2022). Computing Model for Alzheimer Prediction Using Support Vector Machine Classifier. 2022 IEEE International Conference on Current Development in Engineering and Technology (CCET), 1–6. IEEE. https://doi.org/10.1109/CCET56606.2022.10080346
Kaur, I., & Sachdeva, R. (2025). Prediction Models for Early Detection of Alzheimer: Recent Trends and Future Prospects. Archives of Computational Methods in Engineering, 32(6), 3565–3592. https://doi.org/10.1007/s11831-025-10246-3
Loshchilov, I., & Hutter, F. (2019). Decoupled Weight Decay Regularization.
Mirzaei, G., & Adeli, H. (2022). Machine learning techniques for diagnosis of alzheimer disease, mild cognitive disorder, and other types of dementia. Biomedical Signal Processing and Control, 72, 103293. https://doi.org/10.1016/j.bspc.2021.103293
Murugan, S., Venkatesan, C., Sumithra, M. G., Gao, X.-Z., Elakkiya, B., Akila, M., & Manoharan, S. (2021). DEMNET: A Deep Learning Model for Early Diagnosis of Alzheimer Diseases and Dementia From MR Images. IEEE Access, 9, 90319–90329. https://doi.org/10.1109/ACCESS.2021.3090474
Nancy Noella, R. S., & Priyadarshini, J. (2023). Machine learning algorithms for the diagnosis of Alzheimer and Parkinson disease. Journal of Medical Engineering & Technology, 47(1), 35–43. https://doi.org/10.1080/03091902.2022.2097326
Rabie El Kharoua. (2024). Alzheimer’s Disease Dataset. https://doi.org/10.34740/KAGGLE/DSV/8668279
Rajesh Khanna, M. (2023). Multi-level classification of Alzheimer disease using DCNN and ensemble deep learning techniques. Signal, Image and Video Processing, 17(7), 3603–3611. https://doi.org/10.1007/s11760-023-02586-z
Salehi, W., Baglat, P., Gupta, G., Khan, S. B., Almusharraf, A., Alqahtani, A., & Kumar, A. (2023). An Approach to Binary Classification of Alzheimer’s Disease Using LSTM. Bioengineering, 10(8), 950. https://doi.org/10.3390/bioengineering10080950
Singh, S. K., & Chaturvedi, A. (2025). Leveraging Handwriting Dynamics, Explainable AI and Machine Learning for Alzheimer Prediction. https://doi.org/10.1007/978-3-031-81342-9_27
Soladoye, A. A., Aderinto, N., Omodunbi, B. A., Esan, A. O., Adeyanju, I. A., & Olawade, D. B. (2025). Enhancing Alzheimer’s disease prediction using random forest: A novel framework combining backward feature elimination and ant colony optimization. Current Research in Translational Medicine, 73(4), 103526. https://doi.org/10.1016/j.retram.2025.103526
Wukkadada, B., Wankhede, K., Rajesh, S., Ria, C., & Chakraborty, T. (2023). Alzheimer Prediction using Machine Learning Algorithm. 2023 Somaiya International Conference on Technology and Information Management (SICTIM), 39–43. IEEE. https://doi.org/10.1109/SICTIM56495.2023.10104951
Yao, Z., Wang, H., Yan, W., Wang, Z., Zhang, W., Wang, Z., & Zhang, G. (2023). Artificial intelligence-based diagnosis of Alzheimer’s disease with brain MRI images. European Journal of Radiology, 165, 110934. https://doi.org/10.1016/j.ejrad.2023.110934

Details

Primary Language

English

Subjects

Deep Learning

Journal Section

Research Article

Authors

Timur Lale ^*
0000-0002-6958-5057
Türkiye

Publication Date

December 3, 2025

Submission Date

August 27, 2025

Acceptance Date

October 20, 2025

Published in Issue

Year 2025 Volume: 28 Number: 4

DOI

https://doi.org/10.17780/ksujes.1772927

IZ

https://izlik.org/JA67KR68JT

Cite

RIS / Bibtex

APA

Lale, T. (2025). TABULAR TRANSFORMER ARCHITECTURE WITH OPTUNA OPTIMIZATION FOR EARLY DIAGNOSIS OF ALZHEIMER’S DISEASE. Kahramanmaraş Sütçü İmam Üniversitesi Mühendislik Bilimleri Dergisi, 28(4), 2014-2031. https://doi.org/10.17780/ksujes.1772927