New approach for estimating rock slope stability using logistic regression analysis

Yong Hee Lee, Jong Ryeol Kim, Daehyeon Kim, Hee Bog Kang

Research output: Chapter in Book/Report/Conference proceedingChapter

1 Citation (Scopus)

Abstract

Several methods for evaluating the stability of rock slopes have been proposed on the basis of site investigation data. These methods, however, may lead to considerably different results (failure or stable), depending on the subjective judgments associated with the selection of evaluation items and the application of weighting factors. To ensure fair application of the weighting factors, a new approach to evaluating the stability of rock slopes using the binary logistic regression analysis is proposed. Compared with other methods, the new approach allows the analysis of slope stability to be the most precise, with approximately 92% accuracy. This finding suggests that the statistical approach using logistic regression analysis that allows fair application of weight factors is more promising than others that require evaluators' subjective judgments. This finding also suggests that the new approach can be useful to both practitioners and researchers in assessing the rock slope stability.

Original languageEnglish
Title of host publicationGeology and Properties of Earth Materials
Pages99-109
Number of pages11
Edition2016
DOIs
Publication statusPublished - Dec 1 2007

Publication series

NameTransportation Research Record
Number2016
ISSN (Print)0361-1981

ASJC Scopus subject areas

  • Civil and Structural Engineering
  • Mechanical Engineering

Fingerprint Dive into the research topics of 'New approach for estimating rock slope stability using logistic regression analysis'. Together they form a unique fingerprint.

  • Cite this

    Lee, Y. H., Kim, J. R., Kim, D., & Kang, H. B. (2007). New approach for estimating rock slope stability using logistic regression analysis. In Geology and Properties of Earth Materials (2016 ed., pp. 99-109). (Transportation Research Record; No. 2016). https://doi.org/10.3141/2016-11