Atmospheric neutron is the main radioactive particle that induced single event effects and seriously imperils the reliability and safety of the aircraft in aeronautics. The sensitive properties of atmospheric neutron single event effect are usually characterized by cross section. Since the existing neutron particle sources are very limited and the test data points are incomplete, it is difficult to obtain an energy-cross section curve which covers a broad spectrum to forecast single event upset rates. In this study, test data for Hitachi HM628512ALP-7 was extracted by the monitor output neutron from Theodor Svedberg Laboratory (TSL), 2005. The prediction methods of sensitive cross section by Support Vector Machine (SVM) and Neural Network (NN) were creatively compared and analysed for neutron induced Single Event Upset (SEU) of Static Random Access Memory (SRAM). Kernel function is Gaussian radial basis functions was chosen in SVM. The study gives a comparatively integrated neutron energy-cross section trend curve (0-300MeV), which based on the predictive curve of SVM method, and the low energy neutron data from Indiana. The results of neutron SEU cross section has rose slowly with the increase of neutron energy, tending to saturate in the high level of energy. Prediction results almost matched the change law of neutron induced sensitive cross section of single event effect. There was little change in the prediction data of cross-section above 210MeV by NN. Research shows the SEU rate predicted by SVM method is more accurate than NN for component level. Prediction results of SVM method were compared with the data obtained by different particle accelerators and the simulation results by Geant4 to test the accuracy of the predicted results. The fitting curve is combined with the modified atmospheric neutron injection distribution model, and the SEU rate of SRAM is obtained. Research results can be provided as a reference for the failure analysis of board level and equipment level, improve the reliability of avionics equipment finally.