Abstarct: A quantitative structure-property relationship (QSPR) approach was employed to predict activity coefficients at infinite dilution for 58 organic compounds and water at six different temperatures in ionic liquid of [BMPYR][TCM].A large number of descxriptors were calculated using Dragon software and the best calculated descxriptors was selected from 18 classes of Dragon descxriptors by stepwise regression (SR) and genetic algorithm baxsed on partial least square (GA-PLS). 12 descxriptors with SR and 10 descxriptors with GA-PLS methods were selected. The selected variables by these methods were used as input for artificial neural network (ANN) and support vector machine (SVM) to construct of models to predict of activity coefficients at infinite dilution of these compounds. The performance of each model was investigated by test set. The obtained results show the superiority of the SR-ANN than other methods. The mean square error (MSE) and absolute average percent deviation (AAD) for the test sets of SR-ANN, GA-ANN, SR-SVM and GA-SVM were 0.0179 , 2.4306% and 0.0201 , 3.483% and 0.0601, 10.135% and 4.031, 28/874%, respectively