Based on Support Vector Machine (SVM) of fatigue life prediction research

Support vector machine is based on the minimal principle of VC dimension theory and structure risk in the statistics, the model is established through the finite sample, and between its complexity and learning ability to find the best balance, to get the best ability of generalization. The characteristic vector of input space is mapped to high dimension though nonlinear mapping, make nonlinear into a linear problem in the high dimension space, complexity of point set computing can be avoided in the high space.