The Research On Methods Of Supply Chain Performance Measurement On Support Vector Machines

The purpose of Supply Chain Management (SCM) is realizes the Supply Chain (SC) overall efficiency to be most superior, enhancement SC pitch point enterprise's competitive power. Whether it can achieves the purpose, to carry on a measurement to the Supply Chain Performance (SCP) to be very important. The Support Vector Machine (SVM) establishment above the Statistical Learning Theory (SLT) Structure Risk Minimization (SRM) principle, it can solve the small sample, partial minimization etc. The thesis in the introduction Supply Chain Performance Measurement (SCPM) and in the SVM correlation theory foundation, has discussed the SVM in the SCPM application.Firstly, in this thesis, the author in the profound understanding SC&SCM correlation theory foundation systematically studied the SCPM concept, the function and the characteristic. In the SCPM Index of design principle and select method research foundation, the thesis revolved to five aspects of SCM including finance, business process, client service, environmental protection and development. The thesis has analyzed the SCPM indexes constitute and established the SCPM index system and index system quantification. Secondly, we put the theory and method of SVM to apply SCPM in this thesis. Therefore, we have solved three related issues. The fist issue is how to evaluate by classifying. We concatenate the vectors of each two SC to be a "big" vector. We can classify such "big" vectors into three types, namely "better", "equal" and "worse", based on what relation between the two SC is. Thus we can tell the relation between any two SC by classifying the "big" vector concatenate from the vectors of them. The second issue is mufti-class classification algorithms of SVM. The traditional SVM only deal with the binary classification. In this thesis, based on three type's mufti-class classification algorithm, we deal with 3-class classification by one against one method, in which three machines are built to distinguish any two classes respectively. The third issue is how to form absolute evaluations based on the results of classification. To provide absolute evaluations, we adopt a round-robin-like mechanism. In which each SC is compared with every other one, and receive a mark based on the result. Such marks are cumulated to get the final evaluation of the SC. In the last, the thesis confirmed this method validity through the examples.It is a complicated problem to bring SVM into SMPM. Although the author has done some works, but there are many aspects need to be improved. In one word, the author hopes sincerely that this thesis can provide one reliable research method for the SMPM, and at the same time, expand more spaces for the application of SVM.