Evaluation and Analysis of Transfer Learning Models Towards the Prediction of Flood

This study presents a comparative analysis of three transfer learning-based models such as EfficientNet, Vision Transformer (ViT), and ResNetfor predicting pluvial flood. A flood dataset comprising 144,401 records with eight key conditioning variables was collected from Kaggle repository organized by the United States Geological Survey (USGS) and Copernicus Climate Data Store and was further used for the implementation of this study.Additionally, historical rainfall and meteorological data were obtained from the Nigerian Meteorological Agency (NiMet) through their official data request portal.Subsequently, the dataset was pre-processed by cleaning and normalizing, transforming features and augmenting them, and partitioned into training, validation and testing sets. All the models were pretrained on ImageNet weights and trained to learn flood-specific spatial patterns. As the experimental findings indicate, ViT has the best accuracy (93.1%), F1-score (0.925), and AUC-ROC (0.95) that are used to capture long-range spatial dependencies. EfficientNet was more accurate (92.3) and had the highest F1-score of 0.915; however, it took the least amount of time to be trained, which is acceptable in terms of real-time use. ResNet obtained 91.5% accuracy and 0.905 F1-score, showing stable feature acquisition at a modest computational price. The paper shows the success of transfer learning in improving the flood prediction in low-data areas. Generally, ViT should be used in the context of high-accuracy, EfficientNet in the context of computational efficiency, and ResNet in the context of robust and reliable modeling. These results help to justify the creation of AI-based flood early warning systems to enhance urban flood risk management.

Flood Prediction; Transfer Learning; Vision Transformer (ViT); EfficientNet; ResNet; Pluvial Flood

1. Borrohou, S., Fissoune, R., & Badir, H. (2025). Critical role of data transformation in preprocessing: Methods, algorithms, and challenges. In Model and Data Engineering (pp. 108–122). Springer. https://doi.org/10.1007/978-3-031-87719-3_9 [Google Scholar] [Crossref]

2. Bouguerra, H., Hasnaoui, Y., & Tachi, S. E. (2023). Multi-source data fusion for flood prediction using transfer learning. Remote Sensing, 15(3), 678. https://doi.org/10.3390/rs15030678 [Google Scholar] [Crossref]

3. Chamatidis, I., Istrati, D., &Lagaros, N. D. (2024). Vision Transformer for flood detection using satellite images from Sentinel-1 and Sentinel-2. Water, 16(12), 1670. https://doi.org/10.3390/w16121670 [Google Scholar] [Crossref]

4. CHIDI, E. U., UDANOR, C. N., & ANOLIEFO, E. (2024). Exploring the Depths of Visual Understanding: A Comprehensive Review on Real-Time Object of Interest Detection Techniques. Preprints. https://doi.org/10.20944/preprints202402.0583.v1 [Google Scholar] [Crossref]

5. Côté, P.-O., Nikanjam, A., Ahmed, N., Humeniuk, D., &Khomh, F. (2024). Data cleaning and machine learning: A systematic literature review. Automated Software Engineering Journal. https://doi.org/10.48550/arXiv.2310.01765 [Google Scholar] [Crossref]

6. Dosovitskiy, A., Beyer, L., Kolesnikov, A., et al. (2021). An image is worth 16x16 words: Transformers for image recognition at scale. International Conference on Learning Representations. https://arxiv.org/abs/2010.11929 [Google Scholar] [Crossref]

7. Ebere Uzoka Chidi, E Anoliefo, C Udanor, AT Chijindu, LO Nwobodo (2025)” A Blind navigation guide model for obstacle avoidance using distance vision estimation based YOLO-V8n; Journal of the Nigerian Society of Physical Sciences, 2292-229; https://doi.org/10.46481/jnsps.2025.2292 [Google Scholar] [Crossref]

8. Hajji, S., Krimissa, S., Abdelrahman, K., et al. (2025). Enhancing flood prediction through remote sensing, machine learning, and Google Earth Engine. Frontiers in Water, 7. https://doi.org/10.3389/frwa.2025.1514047 [Google Scholar] [Crossref]

9. Harbor M.C, Eneh I.I., Ebere U.C. (2021). Nonlinear dynamic control of autonomous vehicle under slip using improved back-propagation algorithm. International Journal of Research and Innovation in Applied Science (IJRIAS); Vol. 6; Issue 9; https://rsisinternational.org/journals/ijrias/DigitalLibrary/volume-6-issue-9/62-68.pdf [Google Scholar] [Crossref]

10. Hasnaoui, Y., Tachi, S. E., Bouguerra, H., & Yaseen, Z. M. (2025). Transfer learning-based deep learning models for flood and erosion detection in coastal areas. Earth Science Informatics, 18, Article 380. https://doi.org/10.1007/s12145-025-01866-1 [Google Scholar] [Crossref]

11. Hasnaoui, Y., Tachi, S. E., Bouguerra, H., & Yaseen, Z. M. (2025). Transfer learning-based deep learning models for flood and erosion detection in coastal areas. Earth Science Informatics, 18, Article 380. https://doi.org/10.1007/s12145-025-01866-1 [Google Scholar] [Crossref]

12. Hasnaoui, Y., Tachi, S. E., Bouguerra, H., & Yaseen, Z. M. (2025). Transfer learning-based deep learning models for flood and erosion detection in coastal areas. Earth Science Informatics, 18, Article 380. https://doi.org/10.1007/s12145-025-01866-1 [Google Scholar] [Crossref]

13. Kekong P.E, Ajah I.A., Ebere U.C. (2019). Real-time drowsy driver monitoring and detection system using deep learning based behavioural approach. International Journal of Computer Sciences and Engineering 9 (1), 11-21; http://www.ijcseonline.isroset.org/pub_paper/2-IJCSE-08441-18.pdf [Google Scholar] [Crossref]

14. Kimura, N., Yoshinaga, I., Sekijima, K., Azechi, I., & Baba, D. (2020). Convolutional neural network coupled with a transfer-learning approach for time-series flood predictions. Water, 12(1), 96. https://doi.org/10.3390/w12010096 [Google Scholar] [Crossref]

15. Kimura, N., Yoshinaga, I., Sekijima, K., Azechi, I., & Baba, D. (2020). Convolutional neural network coupled with a transfer-learning approach for time-series flood predictions. Water, 12(1), 96. https://doi.org/10.3390/w12010096 [Google Scholar] [Crossref]

16. Lin, K., Xu, Y., & Ran, G. (2023). Comparative study of transfer learning strategies for flood forecasting. Journal of Hydrologic Engineering, 28(2), 04023001. https://doi.org/10.1061/(ASCE)HE.1943-5584.0002156 [Google Scholar] [Crossref]

17. Liu, J., Liu, K., & Wang, M. (2023). A residual neural network integrated with a hydrological model for global flood susceptibility mapping. Remote Sensing, 15(9), 2447. https://doi.org/10.3390/rs15092447 [Google Scholar] [Crossref]

18. Martins, P., Cardoso, F., Váz, P., Silva, J., & Abbasi, M. (2025). Performance and scalability of data cleaning and preprocessing tools: A benchmark on large real-world datasets. Data, 10(5), 68. https://doi.org/10.3390/data10050068 [Google Scholar] [Crossref]

19. Ojeda Avilés, E. E., Olmos-Liceaga, D., & Jung, J.-H. (2025). Stratified sampling algorithms for machine learning methods in solving two-scale partial differential equations. Journal of Scientific Computing, 104, Article 110. https://doi.org/10.1007/s10915-025-03024-7 [Google Scholar] [Crossref]

20. Ran, G., Lin, K., & Xu, Y. (2024). Transfer learning for hydrological modeling: A topographic index-based approach. Hydrological Processes, 38(4), e14876. https://doi.org/10.1002/hyp.14876 [Google Scholar] [Crossref]

21. Sharma, N. K., &Saharia, M. (2025). DeepSARFlood: Rapid and automated SAR-based flood inundation mapping using Vision Transformer-based deep ensembles. Scientific Remote Sensing, 100203. https://doi.org/10.1016/j.srs.2025.100203 [Google Scholar] [Crossref]

22. Shokati, H., Seufferheld, K. D., Fiener, P., & Scholten, T. (2025). Rapid flood mapping from aerial imagery using fine-tuned SAM and ResNet-backboned U-Net. EGUsphere Preprints. https://egusphere.copernicus.org/preprints/2025/egusphere-2025-3146/ [Google Scholar] [Crossref]

23. Tan, M., & Le, Q. V. (2020). EfficientNet: Rethinking model scaling for convolutional neural networks. Proceedings of the 37th International Conference on Machine Learning, 97, 6105–6114. https://proceedings.mlr.press/v97/tan19a.html [Google Scholar] [Crossref]

24. Wang, Y., Li, H., & Zhao, J. (2021). Domain adaptation in hydrological transfer learning: A review and future directions. Water Resources Management, 35(12), 4123–4140. https://doi.org/10.1007/s11269-021-02856-2 [Google Scholar] [Crossref]

25. Xu, Y., Lin, K., Hu, C., Wang, S., Wu, Q., Zhang, L., & Ran, G. (2023). Deep transfer learning based on transformer for flood forecasting in data-sparse basins. Journal of Hydrology, 625, 129956. https://doi.org/10.1016/j.jhydrol.2023.129956 [Google Scholar] [Crossref]

26. Yaseen, Z. M., & Al-Madhhachi, H. (2021). Transfer learning in environmental modeling: A hybrid deep learning approach. Environmental Science and Pollution Research, 28(5), 5678–5691. https://doi.org/10.1007/s11356-020-11045-3 [Google Scholar] [Crossref]

27. Zhang, L., Wang, S., & Wu, Q. (2022). Enhancing LSTM-based flood prediction through transfer learning in data-scarce basins. Environmental Modelling & Software, 150, 105345. https://doi.org/10.1016/j.envsoft.2022.105345 [Google Scholar] [Crossref]

28. Zhao, X., & Chen, J. (2022). Autoencoder-based transfer learning for flood susceptibility mapping. Geocarto International, 37(8), 845–860. https://doi.org/10.1080/10106049.2020.1857234 [Google Scholar] [Crossref]

29. Zhao, X., & Chen, J. (2022). Autoencoder-based transfer learning for flood susceptibility mapping. Geocarto International, 37(8), 845–860. https://doi.org/10.1080/10106049.2020.1857234 [Google Scholar] [Crossref]

30. Zhu, S., Wang, Z., Zhang, W., & Yang, J. (2025). Application of the ResNet-Transformer model for runoff prediction based on multi-source data fusion. Water Resources Management. https://doi.org/10.1007/s11269-025-04241-3 [Google Scholar] [Crossref]

• What the Desert Fathers Teach Data Scientists: Ancient Ascetic Principles for Ethical Machine-Learning Practice
• Comparative Analysis of Some Machine Learning Algorithms for the Classification of Ransomware
• Comparative Performance Analysis of Some Priority Queue Variants in Dijkstra’s Algorithm
• Transfer Learning in Detecting E-Assessment Malpractice from a Proctored Video Recordings.
• Dual-Modal Detection of Parkinson’s Disease: A Clinical Framework and Deep Learning Approach Using NeuroParkNet