A Review on Artificial Neural Networks for Analyzing Unsaturated Shear Strength of Soil

Abstract: Geotechnical engineering deals with soil as a material having uncertain behavior and its properties. Thus the analysis of the properties of soil is complex and time consuming. Shear strength is an important property which governs the parameters such as bearing capacity, stresses in soil etc. The analysis of the data regarding shear strength of unsaturated soils is essential from last two decades. Therefore, this paper is focusing on review on Artificial Neural Networks (ANN) for analyzing unsaturated shear strength. This complex interaction encouraged the application of Artificial Intelligence (AI) software to predict the behaviour of various geotechnical engineering applications. AI has superior predictive ability as compared to traditional methods. This has made AI a popular option in geotechnical engineering applications. ANN is a popular example of AI. The use of ANN is found predominant in analyzing the shear strength of unsaturated soils. In this paper type of soil, software used, input parameters, output parameters and statistical parameter are considered for review on artificial neural networks for analyzing unsaturated shear strength. The purpose of this paper is to review and compare the previous most relevant studies based on certain variables. Overall, ANN can be used for analyzing the shear strength of unsaturated soils.

Keywords: Artificial Intelligence, Artificial Neural Networks, Unsaturated shear strength, Root mean square error.