Decision Tree Approach for Soil Liquefaction Assessment

Amir H Gandomi,David A Roke,Mark M Fridline

doi:10.1155/2013/346285

Amir H Gandomi, David A Roke + Show 1 more

Open Access

https://doi.org/10.1155/2013/346285

Copy DOI

Journal: TheScientificWorldJournal	Publication Date: Jan 1, 2013
Citations: 57	License type: CC BY 3.0

Affiliation: University of Akron

Abstract

In the current study, the performances of some decision tree (DT) techniques are evaluated for postearthquake soil liquefaction assessment. A database containing 620 records of seismic parameters and soil properties is used in this study. Three decision tree techniques are used here in two different ways, considering statistical and engineering points of view, to develop decision rules. The DT results are compared to the logistic regression (LR) model. The results of this study indicate that the DTs not only successfully predict liquefaction but they can also outperform the LR model. The best DT models are interpreted and evaluated based on an engineering point of view.

Highlights

Empirical classification techniques remain a highly researched topic, especially for real-world problems
decision tree (DT) techniques have been successfully applied to several realworld problems, they have been rarely used in engineering, especially geotechnical engineering (e.g., [2])
It is clear that soft computing techniques have been widely used for liquefaction modeling. Statistical methods such as logistic regression (LR) and DT have been only used for liquefaction assessment based on cone penetration test (CPT) data

Summary

Introduction

Empirical classification techniques remain a highly researched topic, especially for real-world problems. CPT provides more accurate and reliable data for liquefaction analysis compared to more conventional soil tests, such as cyclic triaxial and simple shear tests. It can be considered as a very good complement to SPT measurements. It is clear that soft computing techniques have been widely used for liquefaction modeling Statistical methods such as LR and DT have been only used for liquefaction assessment based on CPT data. We have forced the liquefaction assessment DTs to use SPT results as the first variable to corroborate with the engineering point of view Another common problem of most classification techniques is that they are not interpretable. The DT results are further compared with logistic regression (LR) analysis as a classical benchmark

Predictive Modeling Techniques

Liquefaction Modeling

Results and Discussion

Summary and Conclusions