Accurate prediction of the shear strength of unsaturated soils is essential for a cost optimized design of these structures. Based on the effective stress equation proposed by Bishop (1959) for unsaturated soils, shear strength in these soils depends on the parameter χ which is a function of soil suction.

An adaptive learning neural network method is utilized to predict the effective stress parameter, χ, in plane strain condition. The proposed network is a multilayer perceptron network that consists of 6 neurons in the input layer representing air entry value, volumetric water content at residual and saturated conditions, slope of Soil Water Characteristic Curve, net vertical stress and suction. A database prepared from direct shear test results available in the literature are used to train and test the network. The results indicate suitability of the proposed approach for estimating the effective stress parameter of unsaturated soils.