Comparison of Adaptive Neuro Fuzzy Inference System (ANFIS) and Support Vector Machines(SVM) for discharge capacity prediction of a sharp-crested weirs

The aim of this study is to apply different methods to investigation the discharging capacity of a sharpcrested curved plan-form weirs through the Adaptive neuro fuzzy inference system (ANFIS) and Support Vector Machines (SVM) techniques. Subsequently, For training and testing of the proposed equation, experimental data of Kumar et al, have been used and prediction of discharge coefficient through the ANFIS and SVM were compared with equations were proposed with Kumar et al and Zahiri. The result showe that proposed artificial inteligince models have sutable accuracy and also result of superior models is related to total upstream head, spillway height and the angle of curvature of axis curve which demonstrate direct relationship between discharge coefficient and hydraulic properties. Moreover, the performance of ANFIS model is a bit better than SVM technique with relatively low error and high correlation values. Determination coefficient of the proposed equation for discharge coefficient have been calculated as 0.993 for the ANFIS model with Hybrid training method and two point Gaussian membership function, Also this parameter calculated for SVM with RBF Kernel type and with having values include 3, 10 and 0.1 that is related to γ,c and ε respectively as 0.98 for testing phases.