For the characteristics of malfunction diagnose system a model to classify fault printing based on reduced support vector machines (RSVM) is discussed. The printing malfunctions have many classes. There are massive datasets used in fraud detection. The support vector machines have been promising methods for classification because of their solid mathematical foundation. However they are not favored for large- scale because the training complexity ofSVMis highly dependent on the size of data set. This paper use RSVM with an improved nonlinear Kernel to reduced the size of the quadratic program to be solved and simplified the classification of the nonlinear separating surface. Computational results indicate the RSVM has a good efficiency for adjustable printing fault, and computational times as well as memory usage are much smaller for RSVM than that of conventional SVM.