Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models

Hakan Alper Kamiloğlu; Alper Gücüyener

doi:10.18400/tjce.1711510

EN TR

Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models

Öz

In geotechnical engineering, accurately predicting the seismic bearing capacity of shallow foundations on cohesive slopes requires proper consideration of nonlinear effects and parameter interactions. However, most studies in the literature address these complex effects inadequately, either with numerical methods that demand a large number of analyses or with machine learning (ML) models that require large and heterogeneous data sets. In this study, it is aimed to propose an approach that allows for a robust analysis of the system and at the same time provides accurate forecasting for situations where working with small and homogeneous data sets is mandatory. Within the scope of the study, seismic bearing capacity analyses were performed using Plaxis 2D for 273 cases where eight independent variables took different values according to the Face Centered Composite Design (FCCD), and a database was created. Multiple Linear (MLR), Multiple Nonlinear (MNLR) Regression Models were established, and ML models, such as Support Vector Regression (SVR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost), were trained. The models were compared with both in-sample and out-of-sample estimation performance, and parametric sensitivity analyses were evaluated by methods such as ANOVA and SHAP. The results show that the MNLR model with natural logarithmic transformation is the most successful method in terms of both accuracy (R² ≈ 0.98) and reflecting parameter interactions, while the SVR algorithm demonstrates the best generalization ability among ML models under low-data conditions. The results also confirm that the ML models effectively capture the parameter effects.

Anahtar Kelimeler

Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models

Öz

In geotechnical engineering, accurately predicting the seismic bearing capacity of shallow foundations on cohesive slopes requires proper consideration of nonlinear effects and parameter interactions. However, most studies in the literature address these complex effects inadequately, either with numerical methods that demand a large number of analyses or with machine learning (ML) models that require large and heterogeneous data sets. In this study, it is aimed to propose an approach that allows for a robust analysis of the system and at the same time provides accurate forecasting for situations where working with small and homogeneous data sets is mandatory. Within the scope of the study, seismic bearing capacity analyses were performed using Plaxis 2D for 273 cases where eight independent variables took different values according to the Face Centered Composite Design (FCCD), and a database was created. Multiple Linear (MLR), Multiple Nonlinear (MNLR) Regression Models were established, and ML models, such as Support Vector Regression (SVR), Random Forest (RF), and Extreme Gradient Boosting (XGBoost), were trained. The models were compared with both in-sample and out-of-sample estimation performance, and parametric sensitivity analyses were evaluated by methods such as ANOVA and SHAP. The results show that the MNLR model with natural logarithmic transformation is the most successful method in terms of both accuracy (R² ≈ 0.98) and reflecting parameter interactions, while the SVR algorithm demonstrates the best generalization ability among ML models under low-data conditions. The results also confirm that the ML models effectively capture the parameter effects.

Anahtar Kelimeler

Etik Beyan

The submitted work is original and has not been published previously nor is it under consideration for publication elsewhere.

Kaynakça

Cinicioglu O, Erkli A. Seismic bearing capacity of surficial foundations on sloping cohesive ground. Soil Dyn Earthq Eng 2018;111:53–64. https://doi.org/10.1016/j.soildyn.2018.04.027.
Yang G, Xu D, Liu X, Jia Q, Xu H. Ultimate bearing capacity of embedded strip footings on rock slopes under inclined loading conditions using adaptive finite element limit analysis. Sci Rep 2025;15:8641. https://doi.org/10.1038/s41598-025-91234-2.
Zhou H, Zheng G, Yin X, Jia R, Yang X. The bearing capacity and failure mechanism of a vertically loaded strip footing placed on the top of slopes. Comput Geotech 2018;94:12–21. https://doi.org/10.1016/j.compgeo.2017.08.009.
Meyerhof GG. The ultimate bearing capacity of foundation on slopes. Proc. 4th Int. Conf. soil Mech., London: 1957, p. 384–6.
Hansen JB. A revised and extended formula for bearing capacity. DGI Bull, No 28, Danish Geotech Inst 1970:5–11.
Vesic AS. Bearing capacity of shallow foundations. Found. Eng. Hand-b. 1st edn., New York: Van Nostrand Reinhold Co. Inc.; 1975, p. 121–47.
Peynircioğlu H. Test on bearing capacity of shallow foundations on horizontal top surfaces of sand fills. Proc. 2nd Int. Conférence Soil Mech. Found. Eng., Rotterdam: 1948, p. 194.
Castelli F, Motta E. Bearing capacity of strip footings near slopes. Geotech Geol Eng 2010;28:187–98. https://doi.org/https://doi.org/10.1007/s10706-009-9277-9.

Georgiadis K. An upper-bound solution for the undrained bearing capacity of strip footings at the top of a slope. Géotechnique 2010;60:801–6. https://doi.org/10.1680/geot.09.T.016.
Yu Q, Zhang Y, Li D, Chen Q, Lin L, Zhang S, et al. System reliability analysis of bedrock-soil layer slopes subject to seismic action based on the upper bound limit analysis method. Geomatics, Nat Hazards Risk 2025;16. https://doi.org/10.1080/19475705.2024.2442020.
Shirazizadeh S, Keshavarz A, Beygi M, Vali R, Saberian M, Li J, et al. Bearing Capacity of strip footing on Hoek–Brown rock slopes with an underground circular void by lower bound limit analysis. Transp Infrastruct Geotechnol 2025;12:62. https://doi.org/10.1007/s40515-024-00515-2.
Nandi S, Ghosh P. Ultimate bearing capacity of strip footings on Hoek-Brown rock slopes – A stress characteristics solution. Comput Geotech 2024;166:105963. https://doi.org/10.1016/j.compgeo.2023.105963.
Griffiths DV, Martin CM. Critical failure mechanisms in relatively flat undrained slopes. Géotechnique Lett 2020;10:95–9. https://doi.org/10.1680/jgele.19.00075.
Das S, Maheshwari BK. Bearing capacity of strip footings on slopes under eccentric and inclined loads. Geotech Geol Eng 2025;43:93. https://doi.org/10.1007/s10706-024-03053-3.
Cure E, Sadoglu E, Turker E, Uzuner BA. Decrease trends of ultimate loads of eccentrically loaded model strip footings close to a slope. Geomech Eng 2014;6:469–85. https://doi.org/10.12989/gae.2014.6.5.469.
Khalvati Fahliani H, Arvin MR, Hataf N, Khademhosseini A. Experimental model studies on strip footings resting on geocell-reinforced sand slopes. Int J Geosynth Gr Eng 2021;7:24. https://doi.org/10.1007/s40891-021-00270-1.
Bashir K, Jakka RS. Lateral capacity and failure mechanisms of skirted foundation resting on slopes. Acta Geotech 2025;20:89–117. https://doi.org/10.1007/s11440-024-02486-7.
Pham GH, Duong NT, Tran DT, Keawsawasvong S, Lai VQ. Stability analysis of ring foundations on slope crest: 3D FELA and ANN. Transp Infrastruct Geotechnol 2025;12:70. https://doi.org/10.1007/s40515-024-00529-w.
Lai VQ, Shiau J, Van CN, Tran HD, Keawsawasvong S. Bearing capacity of conical footing on anisotropic and heterogeneous clays using FEA and ANN. Mar Georesources Geotechnol 2023;41:1053–70. https://doi.org/10.1080/1064119X.2022.2113485.
Suppakul R, Chavda JT, Jitchaijaroen W, Keawsawasvong S, Rattanadecho P. Soft computing-based models for estimating undrained bearing capacity factor of open caisson in heterogeneous clay. Geotech Geol Eng 2024;42:5335–61. https://doi.org/10.1007/s10706-024-02789-2.
Hanna AM, Nguyen TQ. An axisymmetric model for ultimate capacity of a single pile in sand. Soils Found 2002;42:47–58. https://doi.org/10.3208/sandf.42.2_47.
Lai VQ, Shiau J, Van CN, Tran HD, Keawsawasvong S. Bearing capacity of conical footing on anisotropic and heterogeneous clays using FEA and ANN. Mar Georesources Geotechnol 2023;41:1053–70. https://doi.org/10.1080/1064119X.2022.2113485.
Bashir K, Jakka RS. Lateral capacity and failure mechanisms of skirted foundation resting on slopes. Acta Geotech 2025;20:89–117. https://doi.org/10.1007/s11440-024-02486-7.
Liu X, Jia Q, Galindo RA, Mao F. Probabilistic bearing capacity analysis of square and rectangular footings on cohesive soil slopes considering three-dimensional rotational anisotropy. Comput Geotech 2025;181:107117. https://doi.org/10.1016/j.compgeo.2025.107117.
Luo N, Luo Z. Reliability analysis of embedded strip footings in rotated anisotropic random fields. Comput Geotech 2021;138:104338. https://doi.org/10.1016/j.compgeo.2021.104338.
Luo N, Luo Z. Reliability analysis of embedded strip footings in rotated anisotropic random fields. Comput Geotech 2021;138:104338. https://doi.org/10.1016/j.compgeo.2021.104338.
Halder K, Chakraborty D. Influence of soil spatial variability on the response of strip footing on geocell-reinforced slope. Comput Geotech 2020;122:103533. https://doi.org/10.1016/j.compgeo.2020.103533.
Gu K-Y, Tran NX, Han JM, Kim K-S, Ham K-W, Kim S-R. Semianalytical Solution for the Uplift Bearing Capacity of Spread Foundations in Sand. Int J Geomech 2023;23. https://doi.org/10.1061/IJGNAI.GMENG-8167.
Zięba Z, Krokowska M, Wyjadłowski M, Kozubal JV, Kania T, Mońka J. Assessing the Scale Effect on Bearing Capacity of Undrained Subsoil: Implications for Seismic Resilience of Shallow Foundations. Materials (Basel) 2023;16:5631. https://doi.org/10.3390/ma16165631.
Abdelhamid M, Czekanski A. An experimental investigation of analytical vs. numerical lattice structure design tools. Mech Adv Mater Struct 2022;29:3614–22. https://doi.org/10.1080/15376494.2021.1892887.
Ahmad F, Tang X-W, Ahmad M, Najeh T, Gamil Y. A scientometrics review of conventional and soft computing methods in the slope stability analysis. Front Built Environ 2024;10. https://doi.org/10.3389/fbuil.2024.1373092.
Zhang W, Zhang R, Wu C, Goh ATC, Lacasse S, Liu Z, et al. State-of-the-art review of soft computing applications in underground excavations. Geosci Front 2020;11:1095–106. https://doi.org/10.1016/j.gsf.2019.12.003.
Lai VQ, Lai F, Yang D, Shiau J, Yodsomjai W, Keawsawasvong S. Determining seismic bearing capacity of footings embedded in cohesive soil slopes using multivariate adaptive regression splines. Int J Geosynth Gr Eng 2022;8:46. https://doi.org/10.1007/s40891-022-00390-2.
Mase LZ, Misliniyati R, Muharama NA, Supriani F, Ahmad DA, Fernanda R, et al. Integrating Multiple Linear Regression Analysis and Machine Learning models to predict the bearing capacity of strip footings on sandy clay slopes. Transp Infrastruct Geotechnol 2025;12:90. https://doi.org/10.1007/s40515-025-00544-5.
Li H, Hosseini S, Gordan B, Zhou J, Ullah S. Dimensionless machine learning: Dimensional analysis to improve LSSVM and ANN models and predict bearing capacity of circular foundations. Artif Intell Rev 2025;58:117. https://doi.org/10.1007/s10462-024-11099-1.
Meethal RE, Kodakkal A, Khalil M, Ghantasala A, Obst B, Bletzinger K-U, et al. Finite element method-enhanced neural network for forward and inverse problems. Adv Model Simul Eng Sci 2023;10:6. https://doi.org/10.1186/s40323-023-00243-1.
Pham GH, Duong NT, Tran DT, Keawsawasvong S, Lai VQ. Stability analysis of ring foundations on slope crest: 3D FELA and ANN. Transp Infrastruct Geotechnol 2025;12:70. https://doi.org/10.1007/s40515-024-00529-w.
Liu Y, Liang Y. Integrated machine learning for modeling bearing capacity of shallow foundations. Sci Rep 2024;14:8319. https://doi.org/10.1038/s41598-024-58534-5.
Linardatos P, Papastefanopoulos V, Kotsiantis S. Explainable AI: A Review of Machine Learning Interpretability Methods. Entropy 2020;23:18. https://doi.org/10.3390/e23010018.
Wang H, Liang Q, Hancock JT, Khoshgoftaar TM. Feature selection strategies: a comparative analysis of SHAP-value and importance-based methods. J Big Data 2024;11:44. https://doi.org/10.1186/s40537-024-00905-w.
Hurtado-Alonso N, Manso-Morato J, Revilla-Cuesta V, Skaf M, Ortega-López V. Optimization of cementitious mixes through response surface method: a systematic review. Arch Civ Mech Eng 2024;25:54. https://doi.org/10.1007/s43452-024-01112-3.
Siamardi K. Optimization of fresh and hardened properties of structural light weight self-compacting concrete mix design using response surface methodology. Constr Build Mater 2022;317:125928. https://doi.org/10.1016/j.conbuildmat.2021.125928.
Shi Q, Wang Z, Ke X, Zheng Z, Zhou Z, Wang Z, et al. Trajectory optimization of wall-building robots using response surface and non-dominated sorting genetic algorithm III. Autom Constr 2023;155:105035. https://doi.org/10.1016/j.autcon.2023.105035.
Güllü H, Fedakar Hİ. Response surface methodology for optimization of stabilizer dosage rates of marginal sand stabilized with Sludge Ash and fiber based on UCS performances. KSCE J Civ Eng 2017;21:1717–27. https://doi.org/10.1007/s12205-016-0724-x.
Wang X, Kim S, Wu Y, Liu Y, Liu T, Wang Y. Study on the optimization and performance of GFC soil stabilizer based on response surface methodology in soft soil stabilization. Soils Found 2023;63:101278. https://doi.org/10.1016/j.sandf.2023.101278.
Kamiloğlu HA, Turan H. Estimation of shear strength parameters of a high plasticity clayey soil stabilized with lime at different curing temperatures using Response Surface Methodology (RSM). Pamukkale Üniversitesi Mühendislik Bilim Derg n.d. https://doi.org/10.5505/pajes.2021.83707.
Song L, Yu X, Xu B, Pang R, Zhang Z. 3D slope reliability analysis based on the intelligent response surface methodology. Bull Eng Geol Environ 2021;80:735–49. https://doi.org/10.1007/s10064-020-01940-6.
Hamrouni A, Sbartai B, Dias D. Ultimate dynamic bearing capacity of shallow strip foundations - Reliability analysis using the response surface methodology. Soil Dyn Earthq Eng 2021;144:106690. https://doi.org/10.1016/j.soildyn.2021.106690.
Fathipour H, Javankhoshdel S, Abolfazlzadeh Y, Payan M, Chenari RJ. Probabilistic assessment of seismic bearing capacity of strip footings seated on heterogeneous slopes using Finite Element Limit Analysis (FELA) and Response Surface Method (RSM). Proc. TMIC 2022 Slope Stab. Conf. (TMIC 2022), Dordrecht: Atlantis Press International BV; 2023, p. 199–209. https://doi.org/10.2991/978-94-6463-104-3_18.
Hamrouni A, Sbartai B, Dias D. Probabilistic analysis of ultimate seismic bearing capacity of strip foundations. J Rock Mech Geotech Eng 2018;10:717–24. https://doi.org/10.1016/j.jrmge.2018.01.009.
Ranjbarnia M, Zarei F, Goudarzy M. Probabilistic analysis of bearing capacity of square and strip foundations on rock mass by the Response Surface Methodology. Rock Mech Rock Eng 2023;56:343–62. https://doi.org/10.1007/s00603-022-03090-5.
Dahal BK, Regmi S, Paudyal K, Dahal D, KC D. Enhancing Deep Excavation Optimization: Selection of an Appropriate Constitutive Model. CivilEng 2024;5:785–800. https://doi.org/10.3390/civileng5030041.
Tang Y, Taiebat HA, Russell AR. Bearing Capacity of Shallow Foundations in Unsaturated Soil Considering Hydraulic Hysteresis and Three Drainage Conditions. Int J Geomech 2017;17. https://doi.org/10.1061/(ASCE)GM.1943-5622.0000845.
Kamiloğlu HA. Evaluating the Effects of Reference Pressure Variation in Hardening Soil Model: From Soil Parameters to Finite Element Predictions. In: Prof. Dr. Meltem SARIOĞLU CEBECİ, editor. NEW VISIONS Eng. CONCEPTS - Theor. - Appl., Duvar Yayınları; 2024, p. 52–68.
Das, Braja M., Sivakugan N. Fundamentals of Geotechnical Engineering. 5th editio. Boston, MA: Cengage Learning; 2017.
Craig RF. Craig’s Soil Mechanics. 7th Editio. London and New York: Spon Press, Taylor and Francis Group; 2004.
Keshavarz A, Beygi M, Vali R. Undrained seismic bearing capacity of strip footing placed on homogeneous and heterogeneous soil slopes by finite element limit analysis. Comput Geotech 2019;113:103094. https://doi.org/10.1016/j.compgeo.2019.103094.
Owhorndah, Nanaka S, Iyai D. Exploring the Impact of Key Factors and their Interactions on Process Outcomes in Experimental Optimization. Asian J Probab Stat 2025;27:63–81. https://doi.org/10.9734/ajpas/2025/v27i7777.
Oh WT, Vanapalli SK. Modeling the stress versus settlement behavior of shallow foundations in unsaturated cohesive soils extending the modified total stress approach. Soils Found 2018;58:382–97. https://doi.org/10.1016/j.sandf.2018.02.008.
Richards R, Elms DG, Budhu M. Seismic bearing capacity and settlements of foundations. J Geotech Eng 1993;119:662–74. https://doi.org/10.1061/(ASCE)0733-9410(1993)119:4(662).
Cutler A, Cutler DR, Stevens JR. Random Forests. Ensemble Mach. Learn. Methods Appl., 2012, p. 157–75.
Boser BE, Guyon IM, Vapnik VN. A training algorithm for optimal margin classifiers. COLT ’92 Proc. fifth Annu. Work. Comput. Learn. theory, 1992, p. 144–52.
Kandiri A, Shakor P, Kurda R, Deifalla AF. Modified Artificial Neural Networks and Support Vector Regression to predict lateral pressure exerted by fresh concrete on formwork. Int J Concr Struct Mater 2022;16:64. https://doi.org/10.1186/s40069-022-00554-4.
Kusakabe O, Kimura T, Yamaguchi H. Bearing capacity of slopes under strip loads on the top surfaces. Soils Found 1981;21:29–40. https://doi.org/10.3208/sandf1972.21.4_29.

Ayrıntılar

Birincil Dil

İngilizce

Konular

İnşaat Geoteknik Mühendisliği

Bölüm

Araştırma Makalesi

Yazarlar

Hakan Alper Kamiloğlu ^*
0000-0003-3313-9239
Türkiye

Alper Gücüyener
0009-0001-5300-2355
Türkiye

Erken Görünüm Tarihi

8 Aralık 2025

Yayımlanma Tarihi

4 Mayıs 2026

Gönderilme Tarihi

1 Haziran 2025

Kabul Tarihi

5 Aralık 2025

Yayımlandığı Sayı

Yıl 2026 Cilt: 37 Sayı: 3

DOI

https://doi.org/10.18400/tjce.1711510

IZ

https://izlik.org/JA99XX83FC

Kaynak Göster

RIS / Bibtex

APA

Kamiloğlu, H. A., & Gücüyener, A. (2026). Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models. Turkish Journal of Civil Engineering, 37(3), 47-79. https://doi.org/10.18400/tjce.1711510

AMA

1.Kamiloğlu HA, Gücüyener A. Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models. tjce. 2026;37(3):47-79. doi:10.18400/tjce.1711510

Chicago

Kamiloğlu, Hakan Alper, ve Alper Gücüyener. 2026. “Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models”. Turkish Journal of Civil Engineering 37 (3): 47-79. https://doi.org/10.18400/tjce.1711510.

EndNote

Kamiloğlu HA, Gücüyener A (01 Mayıs 2026) Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models. Turkish Journal of Civil Engineering 37 3 47–79.

IEEE

[1]H. A. Kamiloğlu ve A. Gücüyener, “Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models”, tjce, c. 37, sy 3, ss. 47–79, May. 2026, doi: 10.18400/tjce.1711510.

ISNAD

Kamiloğlu, Hakan Alper - Gücüyener, Alper. “Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models”. Turkish Journal of Civil Engineering 37/3 (01 Mayıs 2026): 47-79. https://doi.org/10.18400/tjce.1711510.

JAMA

1.Kamiloğlu HA, Gücüyener A. Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models. tjce. 2026;37:47–79.

MLA

Kamiloğlu, Hakan Alper, ve Alper Gücüyener. “Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models”. Turkish Journal of Civil Engineering, c. 37, sy 3, Mayıs 2026, ss. 47-79, doi:10.18400/tjce.1711510.

Vancouver

1.Hakan Alper Kamiloğlu, Alper Gücüyener. Low-Data Modeling of Shallow Foundations on Cohesive Slopes: A Comparison between Hybrid FEM–RSM and ML Models. tjce. 01 Mayıs 2026;37(3):47-79. doi:10.18400/tjce.1711510