Predicting Clay Swelling Pressure: A Comparative Analysis of Advanced Symbolic Regression Techniques
Source abstract
Abstract
Swelling pressure is a key geotechnical property that influences the behaviour and stability of engineering structures built on expansive clayey soils. This pressure can be measured directly through laboratory tests or estimated using indirect methods. This paper analyses a dataset of undisturbed clay samples from southeastern Spain using advanced symbolic regression techniques, namely: deep symbolic regression (PhySO), high-performance symbolic regression (PySR), multi-objective symbolic regression (MOSR), and physics-guided symbolic regression (PGSR). These methods provide interpretable results as equations, unlike standard machine learning models. All generated equations showed high performance (R2 > 0.91 and MAE < 23 kPa) and simplicity, making them suitable for practical engineering applications. PySR yielded the best overall metrics (R2 = 0.933, MAE = 20.49 kPa), particularly excelling in high-pressure ranges, while PhySO demonstrated the most balanced performance, especially for low to medium pressures. MOSR minimized edge-case bias, and PGSR, despite lower overall performance, remained competitive. The plasticity index (PI) was identified as the most influential factor in all models, followed by the percentage of fines. The use of undisturbed samples enhanced the reliability of the findings, and the resulting equations enable a flexible estimation of swelling pressure based on commonly available geotechnical parameters.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Soil and Unsaturated Flow · Landslides and related hazards · Drilling and Well Engineering
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
Department of Civil Engineering, University of Alicante, P.O. Box 99, 03080 Alicante, Spain
References (57)
Show source-provided references
1. Al-Rawas, A.A., and Goosen, M.F. (2006). Expansive Soils: Recent Advances in Characterization and Treatment, CRC Press.10.1201/9780203968079 ↗
2. Gourley, C., Newill, D., and Schreiner, H. (2020). Expansive soils: TRL’s research strategy. Engineering Characteristics of Arid Soils, CRC Press.10.1201/9781003077787-28 ↗
3. Expansive soils-the hidden disaster
4. Geotechnical properties of expansive clay shale in Tabuk, Saudi Arabia10.1016/j.jseaes.2004.07.003 ↗
5. Assessments of swelling anisotropy of Ankara clay10.1016/j.enggeo.2008.12.012 ↗
6. The free swell potential of expansive clays stabilized with the shallow bottom ash mixing method10.1016/j.enggeo.2023.107027 ↗
7. Stabilization of expansive soils using chemical additives: A review10.1016/j.jrmge.2022.02.011 ↗
8. Improvement of swelling chlorite soil using sodium silicate alkali activator10.1016/j.asej.2020.10.019 ↗
9. Recent advances in expansive soil stabilization using admixtures: Current challenges and opportunities
10. Utilization of waste materials in the stabilization of expansive pavement subgrade: An extensive review10.1016/j.conbuildmat.2023.132435 ↗
11. Expansive soil remediation using cement kiln dust as stabilizer
12. A rational method for determining the swelling pressure of expansive soils: Paired swell test (PS-test)10.1007/s11440-022-01465-0 ↗
13. Swelling potential and swelling pressure calculation methods: A comprehensive review10.55214/25768484.v9i2.4442 ↗
14. Prediction of swelling pressure of clays10.1061/JSFEAQ.0001218 ↗
15. Swelling characteristics of compacted, expansive soils10.1346/CCMN.1971.0190406 ↗
16. Vijayvergiya, V., and Ghazzaly, O.I. (August, January 30). Prediction of swelling potential for natural clays. Proceedings of the 3rd International Conference on Expansive Soils, Haifa, Israel.
17. Schultz, J.M., and Jafroudi, S. (2023, January 26–29). Correlation of Expansion Index and Atterberg Limits for Expansive Soils. Proceedings of the Geo-Congress 2023, Los Angeles, CA, USA.10.1061/9780784484708.066 ↗
18. Predicting soil swelling potential using soil classification properties10.1007/s10706-023-02525-2 ↗
19. Indirect estimation of swelling pressure of clayey subgrade under pavement structures10.1007/s13369-017-2546-7 ↗
20. The use of feed-forward and cascade-forward neural networks to determine swelling potential of clayey soils10.1016/j.compgeo.2023.105319 ↗
21. Prediction of swelling pressure of soil using artificial intelligence techniques10.1007/s12665-009-0352-6 ↗
22. Comparative Study Between MLR and ANN Techniques to Predict Swelling Pressure of Expansive Clays10.1007/s10706-022-02099-5 ↗
23. Jalal, F.E., Iqbal, M., Khan, W.A., Jamal, A., Onyelowe, K. (2024). ANN-based swarm intelligence for predicting expansive soil swell pressure and compression strength. Sci. Rep., 14.10.1038/s41598-024-65547-7 ↗
24. Prediction of swelling pressure of expansive soil using machine learning methods10.1007/s42107-024-01205-7 ↗
25. Using empirical correlations and artificial neural network to estimate compressibility of low plasticity clays10.1007/s12517-020-06228-3 ↗
26. Discovering causal relations and equations from data10.1016/j.physrep.2023.10.005 ↗
27. Artificial intelligence in physical sciences: Symbolic regression trends and perspectives10.1007/s11831-023-09922-z ↗
28. Genetic programming as a means for programming computers by natural selection10.1007/BF00175355 ↗
29. Mathematical model for approximating shield tunneling-induced surface settlement via multi-gene genetic programming10.1007/s11440-023-01847-y ↗
30. A true triaxial strength criterion for rocks by gene expression programming10.1016/j.jrmge.2023.03.004 ↗
31. Machine learning-based intelligent modeling of hydraulic conductivity of sandy soils considering a wide range of grain sizes10.1016/j.enggeo.2022.106899 ↗
32. Estimation of the undrained shear strength of east Port-Said clay using the genetic programming10.1016/j.asej.2020.02.007 ↗
33. Indirect Estimation of Swelling Pressure of Expansive Soil: GEP versus MEP Modelling10.1155/2023/1827117 ↗
34. Swelling potential reduction of Spanish argillaceous marlstone Facies Tap soil through the addition of crumb rubber particles from scrap tyres
35. Carcedo, F.J.A., Gijón, M.F., Mazo, C.O., and Rodríguez, J.S. (1986). Mapa Previsor de Riesgos por Expansividad de Arcillas en España a Escala 1: 1.000.000, Instituto Geológico y Minero de España.
36. (1995). Análisis Granulométrico de Suelos por Tamizado (Standard No. UNE 103101:1995).
37. (2019). Standard Test Methods for Laboratory Determination of Water (Moisture) Content of Soil and Rock by Mass (Standard No. ASTM D2216-19).
38. (1994). Determinación del Límite Líquido de un Suelo por el Método del Aparato de Casagrande (Standard No. UNE 103103:1994).
39. (2017). Standard Test Methods for Liquid Limit, Plastic Limit and Plasticity index of Soils (Standard No. ASTM D4318-17e1).
40. (1993). Determinación de la Humedad de un Suelo Mediante Secado en Estufa (Standard No. UNE 103300:1993).
41. (2021). Standard Test Methods for Laboratory Determination of Density and Unit Weight of Soil Specimens (Standard No. ASTM D7263-21).
42. (2011). Standard Practice for Classification of Soils for Engineering Purposes (Unified Soil Classification System) (Standard No. ASTM D2487).
43. (1996). Ensayo Para Calcular la Presión de Hinchamiento de un Suelo en Edómetro (Standard No. UNE 103602:1996).
44. (2021). Standard Test Methods for One-Dimensional Swell or Collapse of Soils (Standard No. ASTM D4546-21).
45. Chen, F.H. (2012). Foundations on Expansive Soils, Elsevier.
46. Petersen, B.K., Landajuela, M., Mundhenk, T.N., Santiago, C.P., Kim, S.K., and Kim, J.T. (2021). Deep symbolic regression: Recovering mathematical expressions from data via risk-seeking policy gradients. arXiv.
47. Deep symbolic regression for physics guided by units constraints: Toward the automated discovery of physical laws10.3847/1538-4357/ad014c ↗
48. Cranmer, M. (2023). Interpretable machine learning for science with PySR and SymbolicRegression. jl. arXiv.
49. Real, E., Aggarwal, A., Huang, Y., and Le, Q.V. (February, January 27). Regularized evolution for image classifier architecture search. Proceedings of the AAAI Conference on Artificial Intelligence, Honolulu, HI, USA.
50. Real, E., Liang, C., So, D., and Le, Q. (2020, January 13–18). Automl-zero: Evolving machine learning algorithms from scratch. Proceedings of the International Conference on Machine Learning, Virtual Event.
51. Ferrari, D., Guidetti, V., Wang, Y., and Curcin, V. (2023, January 11–15). Multi-objective symbolic regression to generate data-driven, non-fixed structure and intelligible mortality predictors using ehr: Binary classification methodology and comparison with state-of-the-art. Proceedings of the AMIA Annual Symposium Proceedings, New Orleans, LO, USA.
52. Ferrari, D., Guidetti, V., and Mandreoli, F. (December, January 28). Multi-objective symbolic regression for data-driven scoring system management. Proceedings of the 2022 IEEE International Conference on Data Mining (ICDM), Orlando, FL, USA.10.1109/ICDM54844.2022.00112 ↗
53. A Fast and Elitist Multiobjective Genetic Algorithm: NSGA-II10.1109/4235.996017 ↗
54. Kingma, D., and Ba, J. (2015). Adam: A method for stochastic optimization 3rd International Conference for Learning Representations (San Diego, CA). arXiv.
55. AI Feynman: A physics-inspired method for symbolic regression10.1126/sciadv.aay2631 ↗
56. AI Feynman 2.0: Pareto-optimal symbolic regression exploiting graph modularity
57. On sensitivity estimation for nonlinear mathematical models