A Hybrid Approach Integrating CNN and Heterogeneous Ensemble Learning for AI-Powered Dermatological Diagnosis
Abstract views: 23
/ 
PDF downloads: 18
Keywords:
Dermatological Diagnosis, Skin Diseases, Ensemble Learning, Transfer Learning, CNNsAbstract
Early and accurate diagnosis of dermatological diseases is of great importance for the
effectiveness of clinical decision support systems and the improvement of patient management. However,
traditional approaches based on manual assessment in dermatological diagnostic processes have
significant limitations such as expert dependence, subjective interpretation differences, and time cost. In
this study, a hybrid classification framework combining convolutional neural network (CNN) based
transfer learning models with ensemble learning methods is proposed to enable automatic and reliable
classification from dermatological images. In this study, deep feature extraction was performed using the
publicly available data set and pre-trained CNN models AlexNet, Xception-41, and NASNetLarge. The
resulting high-dimensional feature vectors were classified using six different ensemble learning
algorithms such as Random Forest (RF), Adaptive Boosting (AdaBoost), Categorical Boosting
(CatBoost), Extra Trees (ExtraTrees), Light Gradient Boosting Machine (LightGBM), and Extreme
Gradient Boosting (XGBoost). The performance of the models was comprehensively evaluated using
accuracy, precision, recall, f1-score, and Area Under Curve (AUC) metrics. Experimental results have
shown that the AlexNet+ExtraTrees hybrid approach exhibits superior performance compared to all other
model combinations. This model achieved the most successful results, with accuracy of 77.27% and
precision of 79.23%. The findings reveal that combining CNN-based deep features with ensemble
learning methods enhances the discriminative representation and overall classification performance in
dermatological disease diagnosis tasks. In conclusion, the proposed CNN-based hybrid approach
supported by transfer learning and ensemble learning is considered to offer an effective, scalable, and
clinically adaptable decision support system for the automated classification of skin diseases.
Downloads
References
Almustafa, K. M. (2025). Predictive modeling and optimization in dermatology: Machine learning for skin disease classification. Computers in Biology and Medicine, 189, 109946. https://doi.org/10.1016/j.compbiomed.2025.109946
Pratama, N. R., Setiadi, D. R. I. M., Harkespan, I., & Ojugo, A. A. (2025). Feature fusion with Albumentation for enhancing monkeypox detection using deep learning models. Journal of Computing Theories and Applications, 2(3), 428–441. https://doi.org/10.62411/jcta.12255
Muthamil Sudar, K., Nagaraj, P., Muneeswaran, V., Panda, B., & Bhoi, A. K. (2025). Dermo classify: A dermatologist skin disease detection and classification using DCNN. Research on Biomedical Engineering, 41, 13. https://doi.org/10.1007/s42600-024-00392-1
Gupta, P., Nirmal, J., & Mehendale, N. (2025). A survey on computer vision approaches for automated classification of skin diseases. Multimedia Tools and Applications, 84, 8673–8705. https://doi.org/10.1007/s11042-024-19301-w
Ali, A., Shahbaz, H., & Damaševičius, R. (2025). xCViT: Improved Vision Transformer network with fusion of CNN and Xception for skin disease recognition with explainable AI. Computers, Materials & Continua, 83(1). https://doi.org/10.32604/cmc.2025.059301
Inthiyaz, S., Altahan, B. R., Ahammad, S. H., Rajesh, V., Kalangi, R. R., Smirani, L. K., Hossain, M. A., & Rashed, A. N. Z. (2023). Skin disease detection using deep learning. Advances in Engineering Software, 175, 103361. https://doi.org/10.1016/j.advengsoft.2022.103361
AlEnezi, N. S. A. (2019). A method of skin disease detection using image processing and machine learning. Procedia Computer Science, 163, 85–92. https://doi.org/10.1016/j.procs.2019.12.090
Chen, J., Jiang, Q., Ai, Z., Wei, Q., Xu, S., Hao, B., Lu, Y., Huang, X., & Chen, L. (2025). Pigmented skin disease classification via deep learning with an attention mechanism. Applied Soft Computing, 170, 112571. https://doi.org/10.1016/j.asoc.2024.112571
Alruwaili, M., & Mohamed, M. (2025). An integrated deep learning model with EfficientNet and ResNet for accurate multi-class skin disease classification. Diagnostics, 15(5), 551. https://doi.org/10.3390/diagnostics15050551
Mohan, J., Sivasubramanian, A., Sowmya, V., & Ravi, V. (2025). Enhancing skin disease classification leveraging transformer-based deep learning architectures and explainable AI. Computers in Biology and Medicine, 190, 110007. https://doi.org/10.1016/j.compbiomed.2025.110007
Gairola, A. K., Kumar, V., Sahoo, A. K., Diwakar, M., Singh, P., & Garg, D. (2025). Multi-feature fusion deep network for skin disease diagnosis. Multimedia Tools and Applications, 84, 419–444. https://doi.org/10.1007/s11042-024-18958-7
Rathinam, V., Ramathilagam, A., Valarmathi, K., & Dinesh, E. V. (2025). S3-GHOSTNET: Skin disease detection via Sand Cat Swarm Optimization deep GHOSTNET with facial images. Australian Journal of Electrical and Electronics Engineering. https://doi.org/10.1080/1448837X.2025.2454803
Van Thanh, H., Duy Quang, N., Minh Phuong, T., Kang-Hyun, J., & Van-Dung, H. (2025). A compact version of EfficientNet for skin disease diagnosis application. Neurocomputing, 620, 129166. https://doi.org/10.1016/j.neucom.2024.129166
Ujjwal S. (2025). Skin Disease Classification Dataset [Dataset]. Kaggle. https://www.kaggle.com/datasets/ujjwalsinha01/skin-disease-classification-dataset (Accessed: December 26, 2025).
