In order to study the performance of silicon photoelectric devices, the I-V characteristic modeling of silicon diode based on least square method is studied. Analyze the I-V characteristics of the diode, polynomial regression analysis algorithm is used to fit the nonlinear data set of experimental measurement. The least square method is used to deduce coefficient and construct the matrix, the mathematical model parameters of the I-V characteristics of the reactive silicon diode were obtained. The accuracy of the fitting empirical formula is verified by the error analysis between the predicted value and the actual measured value. The results show that the algorithm and program can be widely used in polynomial regression analysis of nonlinear data sets.