XGBoost-Deep Residual (FedXGB-ResNet) collaborative porosity prediction in a federated learning framework

Abstract

Privacy protection and multimodal fusion represent significant challenges in the context of cross-institutional geological data collaboration. To address these challenges, this paper proposes a collaborative XGBoost-Deep Residual model (FedXGB-ResNet) within a federated learning framework to achieve high-accuracy porosity prediction. Through a heterogeneous federated integration architecture, the framework combines XGBoost’s efficient modeling of structured well-logging parameters (with a feature importance gain of 41.2%) with ResNet’s deep extraction of spatiotemporal features from seismic images (87% overlap in activation maps between training and validation sets). The model achieved R² scores of 0.87 and 0.83 on North Sea and Bakken oilfield datasets, respectively, representing improvements of 12.7% and 9.3% over the traditional FedAvg-XGBoost baseline. The innovative dual privacy protection mechanisms—gradient obfuscation and Paillier encryption—suppressed the membership inference attack success rate to 13.7% and reduced gradient similarity to 0.19 at a noise scale of 1.5 (privacy budget = 0.75), while increasing communication time by only 23%. A dynamic feature distillation mechanism adaptively fuses multimodal features through gated attention units, narrowing the F1-score gap between high-porosity and low-porosity identification to 5%. Experiments demonstrated that the framework reduces the risk of privacy leakage by 84.7% while retaining 93.4% of the performance of a centralized model, offering a balanced solution for collaborative cross-domain geological data analysis in terms of accuracy, privacy, and efficiency.

Keywords

Federated learning / Porosity prediction / XGBoost / Deep residual network / Multimodal fusion / Geological data analysis

Funding

This research was financially supported by Mahasarakham University; the Scientific Research Fund of Institute of Seismology, China Earthquake Administration and National Institute of Natural Hazards, MEM, (No. IS202226322); the 2025 Doctoral Special Support Program Project of Chengdu Jincheng College (NO.2025JCKY(B)0018, 2025JCKY(B)0012); and Mathematics and Finance Research Center Project of Dazhou Social Science Federation Key Research Base (No. SCMF202505).

Conflict of interest

The authors declare they have no competing interests.

Author contributions

Conceptualization: Tianwen Zhao, Cong Pang, Junyan Li, Palakorn Seenoi, Nipada Papukdee. Formal analysis: Tianwen Zhao, Guoqing Chen, Junyan Li, Palakorn Seenoi, Piyapatr Busababodhin. Investigation: Tianwen Zhao, Junyan Li, Nipada Papukdee, Piyapatr Busababodhin. Methodology: Tianwen Zhao, Cong Pang, Guoqing Chen, Piyapatr Busababodhin. Writing–original draft: Tianwen Zhao, Cong Pang, Guoqing Chen, Piyapatr Busababodhin. Writing–review & editing: Tianwen Zhao, Cong Pang, Guoqing Chen, Piyapatr Busababodhin.

Availability of data

Some data used in this study cannot be shared publicly due to collaborative agreement restrictions, but are available from the corresponding author upon reasonable request.

References

[1] Avseth P , Mukerji T , Mavko G. Quantitative Seismic Interpretation: Applying Rock Physics Tools to Reduce Interpretation Risk . Cambridge, UK: Cambridge University Press; 2005. doi: 10.1017/CBO9780511600074

[2] Mavko G , Mukerji T , Dvorkin J. The Rock Physics Handbook: Tools for Seismic Analysis of Porous Media . 2nd ed. Cambridge, UK: Cambridge University Press; 2009. doi: 10.1017/CBO9780511626753

[3] Mukerji T , Avseth P , Mavko G , Takahashi I , González EF. Statistical rock physics: combining rock physics, information theory, and geostatistics to reduce uncertainty in seismic reservoir characterization. Lead Edge. 2001; 20(3): 313-319. doi: 10.1190/1.1438938

[4] Lin Y , Ren P , Chen Z , et al. Meta matrix factorization for federated rating predictions. In: Proceedings of the 43rd International ACM SIGIR Conference on Research and Development in Information Retrieval. Presented at: SIGIR ‘20; July 25—30, 2020; Virtual Event, China. Association for Computing Machinery; 2020: 981-990. doi: 10.1145/3397271.3401081

[5] Chen S , Xue D , Chuai G , Yang Q , Liu Q. FL—QSAR: a federated learning—based QSAR prototype for collaborative drug discovery. Bioinformatics. 2020; 36(22—23): 5492-5498. doi: 10.1093/bioinformatics/btaa1006

[6] Yin F , Lin Z , Kong Q , et al. FedLoc: Federated learning framework for data—driven cooperative localization and location data processing. IEEE Open J Signal Process. 2020; 1: 187-215. doi: 10.1109/OJSP.2020.3036276

[7] Lu MY , Chen RJ , Kong D , et al. Federated learning for computational pathology on gigapixel whole slide images. Med Image Anal. 2022; 76: 102298. doi: 10.1016/j.media.2021.102298

[8] Lu C , Yu Y , Karimireddy SP , Jordan MI , Raskar R. Federated conformal predictors for distributed uncertainty quantification. In: Krause A, Brunskill E, Cho K, Engelhardt B, Sabato S, Scarlett J, eds. Proceedings of the 40th International Conference on Machine Learning. Presented at: ICML ‘23; July 23—29, 2023; Honolulu, HI, USA. PMLR; 2023: 22942-22964. doi: 10.5555/3618408.3619361

[9] Singhal K , Sidahmed H , Garrett Z , et al. Federated reconstruction: Partially local federated learning. Adv Neural Inf Process Syst . 2021; 34: 11220-11232. doi: 10.5555/3540261.3541119

[10] Durga R , Poovammal E. Fled—block: Federated learning ensembled deep learning blockchain model for covid—19 prediction. Front Public Health. 2022; 10: 892499. doi: 10.3389/fpubh.2022.892499

[11] Bugshan N , Khalil I , Rahman MS , Atiquzzaman M , Yi X , Badsha S. Toward trustworthy and privacy—preserving federated deep learning service framework for industrial internet of things. IEEE Trans Ind Inform. 2022; 19(2): 1535-1547. doi: 10.1109/TII.2022.3209200

[12] Hossain I , Puppala S , Talukder S. Collaborative differentially private federated learning framework for the prediction of diabetic retinopathy. In: Proceedings of the 2023 IEEE 2nd International Conference on AI in Cybersecurity (ICAIC). February 7—9, 2023, Houston, TX, USA. IEEE; 2023: 1-6. doi: 10.1109/ICAIC57335.2023.10044122

[13] Almufareh MF , Tariq N , Humayun M , Almas B. A federated learning approach to breast cancer prediction in a collaborative learning framework. Healthcare . 2023; 11(24): 3185. doi: 10.3390/healthcare11243185

[14] Antunes RS , André da Costa C , Küderle A , Yari IA , Eskofier B. Federated learning for healthcare: Systematic review and architecture proposal. ACM Trans Intell Syst Technol. 2022; 13(4): 1-23. doi: 10.1145/3501813

[15] Abaoud M , Almuqrin MA , Khan MF. Advancing federated learning through novel mechanism for privacy preservation in healthcare applications. IEEE Access. 2023; 11: 83562-83579. doi: 10.1109/ACCESS.2023.3301162

[16] Ma Z , Ruhaiyem NIR , Zhang M , et al. A review of federated learning technology and its research progress in healthcare applications. Appl Intell. 2025; 55(10): 765. doi: 10.1007/s10489—025—06627—7

[17] Ding Y , Chen G. Joint prediction model of reservoir parameters based on multimodal transformer graph neural operator physical constraint network. Int Sci Tech Econ Res. 2026; 4(1): 70-89. doi: 10.71451/ISTAER2604

[18] Chaki S , Routray A , Mohanty WK. Well—log and seismic data integration for reservoir characterization: A signal processing and machine—learning perspective. IEEE Signal Process Mag . 2018; 35(2): 72-81. doi: 10.1109/MSP.2017.2776602

[19] He Y , Zhang H , Wu Z , et al. Porosity prediction of tight reservoir rock using well logging data and machine learning. Sci Rep . 2025; 15(1): 13124. doi: 10.1038/s41598—025—95578—7

[20] Zeng X. Cross—border trade fraud detection via integrated heterogeneous graph neural network and XGBoost. Int Sci Tech Econ Res. 2026; 4(1): 47-69. doi: 10.71451/ISTAER2603

[21] Ma D , Yang H , Yang Z , et al. An Intelligent Method for Real—Time Surface Monitoring of Rock Drillability at the Well Bottom Based on Logging and Drilling Data Fusion. Processes. 2025; 13(3): 668. doi: 10.3390/pr13030668

[22] Liu N , Deng W , Li F , Wang Z , Wu H , Gao J. Recent advances on deep residual learning based seismic data reconstruction: an overview. J Geophys Eng. 2025; 22(3): 854-876. doi: 10.1093/jge/gxae011

[23] Agrawal N , Govil H. A deep residual convolutional neural network for mineral classification. Adv Space Res. 2023; 71(8): 3186-3202. doi: 10.1016/j.asr.2022.11.028

[24] Malbašić T , Bojović PD , Bojović Ž , Šuh J , Vujošević D. Hybrid SDN networks: A multi—parameter server load balancing scheme. J Netw Syst Manag. 2022; 30(2): 30. doi: 10.1007/s10922—022—09642—y

[25] Arbaoui M , Brahmia MEA , Rahmoun A , Zghal M. Federated learning survey: A multi—level taxonomy of aggregation techniques, experimental insights, and future frontiers. ACM Trans Intell Syst Technol. 2024; 15(6): 1-69. doi: 10.1145/3678182

[26] Quan MK , Pathirana PN , Wijayasundara M , et al. Federated learning for cyber physical systems: A comprehensive survey. IEEE Commun Surv Tutor. 2026; 28: 3751-3790. doi: 10.1109/COMST.2025.3570288

[27] Yang P , He J. Model multiple heterogeneity via hierarchical multi—latent space learning. In: Proceedings of the 21st ACM SIGKDD International Conference on Knowledge Discovery and Data Mining. Presented at: KDD ‘15; August 10—13, 2015; Sydney, NSW, Australia. Association for Computing Machinery; 2015: 1375-1384. doi: 10.1145/2783258.2783330

[28] Zhang Y , Liang K , Loo BPY. Measuring dynamic accessibility by metro system under travel time uncertainty based on smart card data. J Transp Geogr. 2025; 127: 104294. doi: 10.1016/j.jtrangeo.2025.104294

[29] Xie L , Liu J , Lu S , Chang TH , Shi Q. An efficient learning framework for federated XGBoost using secret sharing and distributed optimization. ACM Trans Intell Syst Technol. 2022; 13(5): 1-28. doi: 10.1145/3523061

[30] Heiyanthuduwage SR , Altas I , Bewong M , Islam MZ , Deho OB. Decision trees in federated learning: current state and future opportunities. IEEE Access. 2024; 12: 127943-127965. doi: 10.1109/ACCESS.2024.3440998

[31] Zhang K , Luo W , Zhong Y , Ma L , Liu W , Li H. Adversarial spatio—temporal learning for video deblurring. IEEE Trans Image Process. 2018; 28(1): 291-301. doi: 10.1109/TIP.2018.2867733

[32] Lv C , Lv X , Wang Z , et al. A focal quotient gradient system method for deep neural network training. Appl Soft Comput. 2025: 113704. doi: 10.1016/j.asoc.2025.113704

[33] Wen Y , Gao T , Li Z , Zhang J , Zhang K , Chen T. All—in—one weather—degraded image restoration via adaptive degradation—aware self—prompting model. IEEE Trans Multimed. 2025. doi: 10.1109/TMM.2025.3535316

[34] Zhang L , Xu J , Sivaraman A , Lazarus JD , Sharma PK , Pandi V. A two—stage differential privacy scheme for federated learning based on edge intelligence. IEEE J Biomed Health Inform. 2023; 28(6): 3349-3360. doi: 10.1109/JBHI.2023.3306425

[35] Wang W , Shen S , Yuan Y. Forecasting short—term export volumes with hybrid models integrating SARIMA with attention—based LSTM. Int Sci Tech Econ Res. 2026; 4(1): 1-22. doi: 10.71451/ISTAER2601

[36] Gao M , Zhou S , Gu W , et al. MMGPT4LF: Leveraging an optimized pre—trained GPT—2 model with multi—modal cross—attention for load forecasting. Appl Energy. 2025; 392: 125965. doi: 10.1016/j.apenergy.2025.125965

[37] Zhao T , Chen G , Pang C , Busababodhin P. Application and Performance Optimization of SLHS—TCN—XGBoost Model in Power Demand Forecasting. CMES Comput Model Eng Sci. 2025; 143(3): 2883-2917. doi: 10.32604/cmes.2025.066442

[38] Zhang J , Dong Y , Frangopol DM , Zhu S , Yang H. Synergistic operation and maintenance enabling lifecycle—aware opportunistic management of offshore wind energy. Appl Energy. 2026; 408: 127424. doi: 10.1016/j.apenergy.2026.127424

[39] Zhao T , Chen G , Suraphee S , Phoophiwfa T , Busababodhin P. A hybrid TCN—XGBoost model for agricultural product market price forecasting. PLoS ONE. 2025; 20(5): e0322496. doi: 10.1371/journal.pone.0322496