Comparison of artificial neural network and coupled simulated annealing based least square support vector regression models for prediction of compressive strength of high-performance concrete

High-performance concrete (hpc) is a complex composite material with highly nonlinear mechanical behavior. concrete compressive strength, as one of the most essential qualities of concrete, is also a highly nonlinear function of ingredients. in this paper, least square support vector regression (lssvr) model based on coupled simulated annealing (csa) has been successfully used to find the nonlinear relationship between the concrete compressive strength and eight input factors (the cement, the blast furnace slags, the fly ashes, the water, the superplasticizer, the coarse aggregates, the fine aggregates, age of testing). to evaluate the performance of the csa-lssvr model, the results of the hybrid model were compared with those obtained by artificial neural network (ann) model. a comparison study is made using the coefficient of determination r^2 and root mean squared error (rmse) as evaluation criteria. the accuracy, the computational time, the advantages and shortcomings of these modeling methods are also discussed. the training and testing results have shown that anns and csa-lssvr models have strong potential for predicting the compressive strength of hpc.