Journal of Structural and Construction Engineering

Journal of Structural and Construction Engineering

Predicting the probability of liquefaction using Taguchi-optimized artificial neural network and comparing it with the results of different machine learning algorithms

Document Type : Original Article

Authors
Esmaeil Eksiri 1
Mojtaba Karimaei Tabarestani 2
Ali Nabizadeh 3
Mostafa Adresi 2
1 Graduated in Geotechnical Engineering, Faculty of Civil Engineering, Shahid Rajaee Teacher Training University, Tehran, Iran
2 Associate Professor, Faculty of Civil Engineering, Shahid Rajaee Teacher Training University, Tehran, Iran
3 Assistant Professor, Faculty of Civil Engineering, Shahid Rajaee Teacher Training University, Tehran, Iran
10.22065/jsce.2025.523786.3730
Abstract
Soil liquefaction is a phenomenon that occurs due to the loss of resistance of saturated soil mass during an earthquake. In this case, the structure built on this soil loses its static equilibrium and overturns. Therefore, predicting liquefaction is a very important issue in civil engineering. There are various methods for determining the potential of soil liquefaction, most of which are based on field tests and empirical equations, which can sometimes be accompanied by a large approximation. Today, the use of modern artificial intelligence methods is of great importance for solving complex engineering problems. Accordingly, in this paper, an attempt was made to predict the occurrence of liquefaction in a region by using an artificial neural network and optimizing it with the Taguchi design of experiments method. In the first step, by collecting existing data on the occurrence of liquefaction in different parts of the world, two datasets were created based on CPT and SPT tests. Next, using the Taguchi design of experiments method, the components of the artificial neural network, including the number of neurons in the middle layers and the activation function, were optimized, and the highest accuracy was obtained for the two datasets of CPT and SPT, equal to 0.89 and 0.73, respectively. Then, the evaluation of the obtained results with the results of 5 different machine learning models, including support vector machine, logistic fitting, k-nearest neighbor, decision tree, and random forest, showed that the optimized artificial neural network gives the best results. Finally, comparing the results of the present model with the experimental method of calculating the liquefaction coefficient shows the higher accuracy of the present model.
Keywords
Liquefaction
Artificial neural network
Taguchi design of experiments
Datasets
Optimization
Subjects
[1] Alavi, A.H. and Gandomi, A. H. (2011). Energy-based numerical models for assessment of soil liquefaction. Geosci. Front. 3(4), 541–555.
[2] Farrokhzad, F., Choobbasti, A. J. and Barari, A. (2012). Liquefaction microzonation of Babol city using artificial neural network, J. King Saud Univ. – Sci. 24(1), 89–100.
[3] Youd et al. (2001). Liquefaction Resistance of Soils: Summary Report from the 1996 NCEER and 1998 NCEER/NSF Workshops on Evaluation of Liquefaction Resistance of Soils. J. Geotech. & Geo-environ. Eng. ASCE, 127(10), 817-833.
[4] Hwang, J. H. and Yang, C. W. (2001). Verification of critical cyclic strength curve by Taiwan Chi-Chi earthquake data. Soil Dyn. Earthq. Eng., 21(3), 237–257.
[5] Cetin K. O. (2004). Standard Penetration Test-Based Probabilistic and Deterministic Assessment of Seismic Soil Liquefaction Potential. J. Geotech. Geoenvironmental Eng. 130(12), 1314–1340.
[6] Muduli P. K. and Das, S. K. (2014). Evaluation of liquefaction potential of soil based on standard penetration test using multi-gene genetic programming model, Acta Geophys. 62(3), 529–543.
[7] Juang, C. H.  Yuan, H. Lee, D. H. and Lin, P.S. (2003). Simplified Cone Penetration Test-based Method for Evaluating Liquefaction Resistance of Soils. J. Geotech. Geoenvironmental Eng. 129(1), 66–80.
[8] Pal, M. (2006). Support vector machines-based modelling of seismic liquefaction potential. Int. J. Numer. Anal. Methods Geomech. 30(10), 983–996.
[9] Hanna, A. M.  Ural, D. and Saygili, G.  (2007). Neural network model for liquefaction potential in soil deposits using Turkey and Taiwan earthquake data. Soil Dyn. Earthq. Eng. 27(6), 521–540.
[10] Samui P. and Sitharam, T. G. (2011). Machine learning modelling for predicting soil liquefaction susceptibility. Nat. Hazards Earth Syst. Sci. 11(1), 1–9.
[11] Samui P. and Center, J. K.  (2012). Directional response of a reconstituted fine-grained soil - Part II : Performance of different constitutive models. Int. J. Numer. Anal. Methods Geomech. 30(13), 1303–1336.
[12] Xue X. and Yang, X.  (2013). Application of the adaptive neuro-fuzzy inference system for prediction of soil liquefaction. Nat. Hazards, 67(2), 901–917.
[13] Venkatesh, K., Kumar, V. and Tiwari, R. P.  (2013). Appraisal of liquefaction potential using neural network and neuro fuzzy approach, Appl. Artif. Intell. 27(8), 700–720.
[14] Muduli, P. K., Das, S. K. and Bhattacharya, S. (2014). CPT-based probabilistic evaluation of seismic soil liquefaction potential using multi-gene genetic programming. Georisk.  8(1), 14–28.
[15] Kohestani, V. R., Hassanlourad, M.  andArdakani, A.  (2015). Evaluation of liquefaction potential based on CPT data using random forest. Nat. Hazards. 79(2), 1079–1089.
[16] Abbaszadeh Shahri, A. (2016). Assessment and Prediction of Liquefaction Potential Using Different Artificial Neural Network Models: A Case Study. Geotech. Geol. Eng. 34(3), 807–815.
[17] Hu, J. L., Tang, X. W.  and Qiu, J. N. (2016). Assessment of seismic liquefaction potential based on Bayesian network constructed from domain knowledge and history data. Soil Dyn. Earthq. Eng. 89, 49–60.
[18] Xue X. and Liu, E. (2017). Seismic liquefaction potential assessed by neural networks. Environ. Earth Sci. 76(5).
[19] Hoang N. D. and Bui, D. T. (2018). Predicting earthquake-induced soil liquefaction based on a hybridization of kernel Fisher discriminant analysis and a least squares support vector machine: a multi-dataset study. Bull. Eng. Geol. Environ., 77(1), 191–204.
[20] Hu J. and Liu, H. (2019). Bayesian network models for probabilistic evaluation of earthquake-induced liquefaction based on CPT and Vs databases. Eng. Geol. 254, 76–88.
[21] Zhou, J.  Li, E. Wang, M. Chen, X.  Shi, X.  and Jiang, L. (2019). Feasibility of Stochastic Gradient Boosting Approach for Evaluating Seismic Liquefaction Potential Based on SPT and CPT Case Histories. J. Perform. Constr. Facil. 33(3), 1–10.
[22] Atangana Njock, P. G. Shen, S. L. Zhou, A. and Lyu, H. M. (2019).  Evaluation of soil liquefaction using AI technology incorporating a coupled ENN / t-SNE model. Soil Dyn. Earthq. Eng. 130,105988.
[23] Kamura A. et al., (2021). Exploring the possibility of assessing the damage degree of liquefaction based only on seismic records by artificial neural networks. Soils Found. 61(3), 658–674.
[24] Zhou, J. Huang, S. Wang, M. and Qiu, Y. (2021). Performance evaluation of hybrid GA – SVM and GWO – SVM models to predict earthquake ‑ induced liquefaction potential of soil : a multi ‑ dataset investigation. Eng. Comput. 38(3), 4197-4225.
[25] Eric R. and Moss, S. (2003). Cpt-Based Probabilistic Assessment of. MASTER, UNIVERSITY OF CALIFORNIA, BERKELEY.
[26] Goh, A. T. C and Goh, S. H. (2007). Support vector machines: Their use in geotechnical engineering as illustrated using seismic liquefaction data, Comput. Geotech. 34(5), 410–421.
[27] Sharma, S. Activation functionsin Neural Networks. 2022 https://towardsdatascience.com.-activate-function-neural-networks-(accessed Dec. 03, 2022).
[28] Support Vector Machine Algorithm. https://www.javatpoint.com/machine-learning-support-vector
-machine-algorithm(accessed May. 07, 2022).
[29] Logistic Regression in Machine Learning. 2022. https://www.javatpoint.com logistic-regression-in mmachine-learning(accessed May. 06, 2022).
[30] Decision Tree Classification Algorithm. 2022 https://www.javatpoint.com.com/machine-learning-decision-tree-classification-algorithm(accessed Apr. 03, 2022).
[31] K-Nearest Neighbor.2021, https://www.javatpoint.com/k-nearest-neighbor-algorithm-for-machine-learning(accessed Feb. 04, 2022).
[32] Tavish Srivastava, (2023). A Complete Guide to K-Nearest Neighbors,”.https://www.analyticsvidhya. .com/blog/2018/03/introduction-k-neighbours-algorithm-clustering/(accessed Mar. 02, 2023).
[33] Piyush Kashyap, (2024). Understanding Precision, Recall, and F1 Score Metrics,”. https://medium.com /@piyushkashyap045/understanding-precision-recall-and-f1-score-metrics-ea219b908093/(accessed Dec. 02, 2024).
[34] Esfandmaz, S., Feizi, A., Karimaei Tabarestani, M., and Rasi Nezami S. (2021) An artificial neural network and taguchi method integrated approach to predicting the local scour depth around the bridge pier during flood event. J. Hydraul. Struct. 7 (1), 98-113.
[35] Fraley, S. Zalewski, J., Oom, M. and Terrien, B. (2022). Design of Experiments via TaguchiMethods.
https://eng.libretexts.org/Bookshelves/Industrial_and_Systems_Engineering/Chemical_Process_Dynamics_and_Controls_(Woolf). (accessed JUN. 02, 2022).
[36] Goharzay, M. Noorzad, A.  Ardakani, A. M. and Jalal, M.  (2017). A worldwide SPT-based soil liquefaction triggering analysis utilizing gene expression programming and Bayesian probabilistic method. J. Rock Mech. Geotech. Eng. 9(4), 683–693.

  • Receive Date 26 May 2025
  • Revise Date 01 September 2025
  • Accept Date 27 September 2025