Welcome to the InfoSci Platform
MARS and Neural Network Models for Shear Strength Prediction of Squat Reinforced Concrete Walls: Shear Strength of Squat Reinforced Concrete Walls

MARS and Neural Network Models for Shear Strength Prediction of Squat Reinforced Concrete Walls: Shear Strength of Squat Reinforced Concrete Walls

Anthony T.C. Goh, Wengang Zhang
Copyright: © 2017 |Pages: 34
ISBN13: 9781522505884|ISBN10: 1522505881|EISBN13: 9781522505891
DOI: 10.4018/978-1-5225-0588-4.ch010
Cite Chapter Cite Chapter

MLA

Goh, Anthony T.C., and Wengang Zhang. "MARS and Neural Network Models for Shear Strength Prediction of Squat Reinforced Concrete Walls: Shear Strength of Squat Reinforced Concrete Walls." Modeling and Simulation Techniques in Structural Engineering, edited by Pijush Samui, et al., IGI Global, 2017, pp. 294-327. https://doi.org/10.4018/978-1-5225-0588-4.ch010

APA

Goh, A. T. & Zhang, W. (2017). MARS and Neural Network Models for Shear Strength Prediction of Squat Reinforced Concrete Walls: Shear Strength of Squat Reinforced Concrete Walls. In P. Samui, S. Chakraborty, & D. Kim (Eds.), Modeling and Simulation Techniques in Structural Engineering (pp. 294-327). IGI Global. https://doi.org/10.4018/978-1-5225-0588-4.ch010

Chicago

Goh, Anthony T.C., and Wengang Zhang. "MARS and Neural Network Models for Shear Strength Prediction of Squat Reinforced Concrete Walls: Shear Strength of Squat Reinforced Concrete Walls." In Modeling and Simulation Techniques in Structural Engineering, edited by Pijush Samui, Subrata Chakraborty, and Dookie Kim, 294-327. Hershey, PA: IGI Global, 2017. https://doi.org/10.4018/978-1-5225-0588-4.ch010

Export Reference

Mendeley
Favorite

Abstract

Squat walls are widely used in structural engineering. The empirical equations currently used to calculate the peak shear strength of squat walls do not correlate well with the experimental results. Another limitation is the reliance on the use of many assumed intermediate parameters. This chapter explores the use of MARS and BPNN approaches to build predictive peak shear strength models of squat walls based on an extensive experimental database from the literature. First the MARS methodology and its associated procedures will be explained in detail. Analyses of the database are then carried out to verify the MARS's capacity in modelling the non-linear interactions between variables without making any specific assumptions. The performances of the built MARS and BPNN models are compared in terms of predictive accuracy, parameter relative importance, parametric analysis and model interpretability. Design charts are also proposed based on parametric studies using the developed models.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.