Evolusi Performa Arsitektur Deep Learning melalui Optimasi Bertahap dan Interpretabilitas Grad-CAM untuk Klasifikasi Penyakit Ikan Air Tawar

Sasa Kirana Wulandari; Fachruddin Fachruddin; Jasmir Jasmir

doi:10.61132/prosemnasproit.v2i2.179

Authors

Sasa Kirana Wulandari Universitas Dinamika Bangsa https://orcid.org/0009-0001-3593-4327
Fachruddin Fachruddin Universitas Dinamika Bangsa https://orcid.org/0000-0003-3258-2060
Jasmir Jasmir Universitas Dinamika Bangsa https://orcid.org/0000-0003-0059-2512

DOI:

https://doi.org/10.61132/prosemnasproit.v2i2.179

Keywords:

Deep Learning, Fish Disease Classification, GradCAM, Model Optimization, Transfer Learning

Abstract

Freshwater fish diseases significantly affect aquaculture productivity and economic sustainability, while accurate visual classification remains challenging due to interclass similarity and image variability. This study presents a comparative evaluation of three deep learning architectures—DenseNet201, ResNet50, and EfficientNetV2-S—using a stepwise optimization strategy combined with Gradient-weighted Class Activation Mapping (Grad-CAM) for freshwater fish disease classification. Models were trained through three phases: baseline, optimized, and fine-tuned. Performance was evaluated using accuracy, precision, recall, F1 score, Matthews correlation coefficient (MCC), Cohen’s kappa, and per-class ROC–AUC. Results show consistent performance improvement across all architectures, with EfficientNetV2-S achieving the highest accuracy (97.14%), followed by ResNet50 (96.11%) and DenseNet201 (94.40%). High ROC–AUC values (>0.98) indicate strong discriminative capability. Grad-CAM analysis confirms that all optimized models focus on biologically relevant lesion regions, enhancing model transparency and reliability.

References

Ambar, M. K., Oztoprak, H., & Yurtkan, K. (2025). Optimizing ResNet-50 for multiclass classification: A Multi-Stage Learning Approach. IEEE Access.

Archana, R., & Jeevaraj, P. S. E. (2024). Deep learning models for digital image processing: a review. Artificial Intelligence Review, 57(1), 11.

Barulina, M., Okunkov, S., Ulitin, I., & Sanbaev, A. (2023). Sensitivity of modern deep learning neural networks to unbalanced datasets in multiclass classification problems. Applied Sciences, 13(15), 8614.

Bernery, C., Bellard, C., Courchamp, F., Brosse, S., Gozlan, R. E., Jarić, I., Teletchea, F., & Leroy, B. (2022). Freshwater fish invasions: A comprehensive review. Annual Review of Ecology, Evolution, and Systematics, 53(1), 427–456.

Biswas, S., Muduli, D., Islam, M. A., Kanade, A. S., Zamani, A. T., Kanade, S. P., & Parveen, N. (2024). Empirical Evaluation of Deep Learning Techniques for Fish Disease Detection in Aquaculture Systems: A Transfer Learning and Fusion-Based Approach. IEEE Access, 12, 176136–176154. https://doi.org/10.1109/ACCESS.2024.3504283

Canbek, G., Taskaya Temizel, T., & Sagiroglu, S. (2022). PToPI: A comprehensive review, analysis, and knowledge representation of binary classification performance measures/metrics. SN Computer Science, 4(1), 13.

Cardoso, P. H. M., Moreno, L. Z., De Oliveira, C. H., Gomes, V. T. M., Silva, A. P. S., Barbosa, M. R. F., Sato, M. I. Z., Balian, S. C., & Moreno, A. M. (2021). Main bacterial species causing clinical disease in ornamental freshwater fish in Brazil. Folia Microbiologica, 66(2), 231–239.

Das, S. K., & Mandal, A. (2022). Diversification in aquaculture resources and practices for smallholder farmers. In Agriculture, Livestock Production and Aquaculture: Advances for Smallholder Farming Systems Volume 1 (pp. 263–286). Springer.

Diallo, R., Edalo, C., & Awe, O. O. (2024). Machine learning evaluation of imbalanced health data: a comparative analysis of balanced accuracy, MCC, and F1 score. In Practical Statistical Learning and Data Science Methods: Case Studies from LISA 2020 Global Network, USA (pp. 283–312). Springer.

Ergün, E. (2025). SwinFishNet: A Swin Transformer-based approach for automatic fish species classification using transfer learning. PLoS One, 20(5), e0322711.

Ghnemat, R., Alodibat, S., & Abu Al-Haija, Q. (2023). Explainable artificial intelligence (XAI) for deep learning based medical imaging classification. Journal of Imaging, 9(9), 177.

Gholizade, M., Soltanizadeh, H., Rahmanimanesh, M., & Sana, S. S. (2025). A review of recent advances and strategies in transfer learning. International Journal of System Assurance Engineering and Management, 1–40.

Giudici, P., Rosciano, R. C., Schrader, J., & Kummerfeld, D.-M. (2025). A Multiclass ROC Curve. ArXiv Preprint ArXiv:2510.15670.

Giuffrida, R., de Majo, C., Giuffrida, M., & Broadbent, I. D. (2025). FAIR data management practices to introduce circular economy in aquaculture: benefits, barriers and a preliminary roadmap. International Journal of Data Science and Analytics, 1–21.

Guluwadi, S. (2024). Enhancing brain tumor detection in MRI images through explainable AI using Grad-CAM with Resnet 50. BMC Medical Imaging, 24(1), 1–19.

Haddad, M. H., & Hassan Mohammed, F. (2025). A CONVOLUTIONAL NEURAL NETWORK APPROACH FOR PRECISION FISH DISEASE DETECTION. https://doi.org/10.46121/pspc.53.2.23

Hariri, M., Avsar, E., & Aydın, A. (2025). LatentResNet: An Optimized Underwater Fish Classification Model with a Low Computational Cost. Journal of Marine Science and Engineering, 13(6), 1019.

Hasan, N., Ibrahim, S., & Aqilah Azlan, A. (2022). Fish diseases detection using convolutional neural network (CNN). International Journal of Nonlinear Analysis and Applications, 13(1), 1977–1984.

Huang, Y.-P., & Khabusi, S. P. (2023). A CNN-OSELM multi-layer fusion network with attention mechanism for fish disease recognition in aquaculture. IEEe Access, 11, 58729–58744.

Ibrahim, A., Zayed, T., & Lafhaj, Z. (2024). Enhancing construction performance: A critical review of performance measurement practices at the project level. Buildings, 14(7), 1988.

Isaac, S., Daniel, O., Grivin, M., Kyagaba, J. M., Lule, E., & Marvin, G. (2024). Explainable Machine Vision Techniques for Fish Disease Detection with Deep Transfer Learning. 2024 5th International Conference on Electronics and Sustainable Communication Systems (ICESC), 1218–1227. https://doi.org/10.1109/ICESC60852.2024.10689742

Kalusivalingam, A. K., Sharma, A., Patel, N., & Singh, V. (2021). Enhancing diagnostic accuracy in medical imaging through convolutional neural networks and transfer learning algorithms. International Journal of AI and ML, 2(3).

Liao, L., Li, H., Shang, W., & Ma, L. (2022). An empirical study of the impact of hyperparameter tuning and model optimization on the performance properties of deep neural networks. ACM Transactions on Software Engineering and Methodology (TOSEM), 31(3), 1–40.

MacAulay, S., Ellison, A. R., Kille, P., & Cable, J. (2022). Moving towards improved surveillance and earlier diagnosis of aquatic pathogens: From traditional methods to emerging technologies. Reviews in Aquaculture, 14(4), 1813–1829.

Mamun, M. R. I., Rahman, U. S., Akter, T., & Azim, M. A. (2023). Fish Disease Detection using Deep Learning and Machine Learning. International Journal of Computer Applications, 975, 8887.

Mandal, A., & Ghosh, A. R. (2024). Role of artificial intelligence (AI) in fish growth and health status monitoring: A review on sustainable aquaculture. Aquaculture International, 32(3), 2791–2820.

Minh, D., Wang, H. X., Li, Y. F., & Nguyen, T. N. (2022). Explainable artificial intelligence: a comprehensive review. Artificial Intelligence Review, 55(5), 3503–3568.

Mohanty, S., & Sahoo, P. K. (2025). Role of Environment in Aquatic Animal Health and Noninfectious Diseases in Finfish Aquaculture. In Aquatic Animal Health Management (pp. 475–505). Springer.

Nishat, A. (2024). Regularization Frameworks for Overfitting Prevention in Language Model Fine-Tuning. International Journal of Digital Innovation, 5(1).

Owusu-Adjei, M., Ben Hayfron-Acquah, J., Frimpong, T., & Abdul-Salaam, G. (2023). Imbalanced class distribution and performance evaluation metrics: A systematic review of prediction accuracy for determining model performance in healthcare systems. PLOS Digital Health, 2(11), e0000290.

Rimmer, M. A., Larson, S., Lapong, I., Purnomo, A. H., Pong-Masak, P. R., Swanepoel, L., & Paul, N. A. (2021). Seaweed aquaculture in Indonesia contributes to social and economic aspects of livelihoods and community wellbeing. Sustainability, 13(19), 10946.

Roscher, M. B., Allison, E. H., Mills, D. J., Eriksson, H., Hellebrandt, D., & Andrew, N. L. (2022). Sustainable development outcomes of livelihood diversification in small‐scale fisheries. Fish and Fisheries, 23(4), 910–925.

Shubham, S., Jain, N., Gupta, V., Mohan, S., Ariffin, M. M., & Ahmadian, A. (2023). Identify glomeruli in human kidney tissue images using a deep learning approach. Soft Computing, 27(5), 2705–2716. https://doi.org/10.1007/s00500-021-06143-z

Sultana, M., Bhattacharya, S., Adhikari, N., Sengupta, D., & De, D. (2024). Artificial intelligence for sustainable ocean health. In Artificial intelligence and edge computing for sustainable ocean health (pp. 3–38). Springer.

Suzuki, A. (2021). Rising importance of aquaculture in Asia: current status, issues, and recommendations. Asian Development Outlook Update Background Paper. Asian Development Bank, Manila.

Szandała, T. (2023). Unlocking the black box of CNNs: Visualising the decision-making process with PRISM. Information Sciences, 642, 119162.

Tang, M., Ting, K. C., & Rashidi, N. H. (2024). DenseNet201-based waste material classification using transfer learning approach. Applied Mathematics and Computational Intelligence (AMCI), 13(2), 113–120.

Teng, Q., Liu, Z., Song, Y., Han, K., & Lu, Y. (2022). A survey on the interpretability of deep learning in medical diagnosis. Multimedia Systems, 28(6), 2335–2355.

Tezzo, X., Bush, S. R., Oosterveer, P., & Belton, B. (2021). Food system perspective on fisheries and aquaculture development in Asia. Agriculture and Human Values, 38(1), 73–90.

Virginia, V. V, Mbanaso, U. M., Chaku, E. S., Clementina, I. O., & Aimufua, G. (2025). Deep learning and convolutional neural networks (CNNs). GAS Journal of Engineering and Technology (GASJET), 2(11), 73–84.

Wulandari, S. K., & Jasmir, J. (2024). Penggunaan Resnet-50 Untuk Deteksi Penyakit Ikan Air Tawar di Akuakultur Studi Kasus pada Akuakultur Asia Selatan. Prosiding Seminar Nasional Bisnis, Teknologi Dan Kesehatan (SENABISTEKES), 1(1), 17–24.

Yang, Y., Theisen, R., Hodgkinson, L., Gonzalez, J. E., Ramchandran, K., Martin, C. H., & Mahoney, M. W. (2023). Test accuracy vs. generalization gap: Model selection in nlp without accessing training or testing data. Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 3011–3021.

Zhang, C., Bengio, S., Hardt, M., Recht, B., & Vinyals, O. (2021). Understanding deep learning (still) requires rethinking generalization. Communications of the ACM, 64(3), 107–115.

Zhao, C. H., & Leong, W. Y. (2025). Optimization solutions and simple innovative solution research on ResNet50 model. ASM Science Journal, 20(1).