Prediction of the Hardness of Cu-3Ti-1Cr Alloy Using Least Square Support Vector Machines

Abstract

A new model based on least square support vector machines (LSSVM) and capable of forecasting the hardness of Cu-3Ti-1Cr alloy has been proposed. Data mining and artificial intelligence techniques were used to examine the forecasting capability of the model. In order to improve predictive accuracy and generalization ability of LSSVM model, leave-one-out-cross-validation (LOOCV) technique was adopted to determine the optimal hyper-parameters of LSSVM automatically. The forecasting performance of the LSSVM model and the partial least squares (PLS) regression integrated with radial basis function (RBF-PLS) has been compared with the experimental values. The result demonstrates that the LSSVM model is superior to the conventional RBF-PLS model in predicting the hardness of Cu-3Ti-1Cr alloy and of better generalization performance than the RBF-PLS model. The present calculated results are consistent with the experimental values. We would expect the proposed LSSVM model as a powerful tool to forecast the variation of the hardness of copper alloys with prior cold work, aging temperature and aging time.