An Improved P-SVM Model for Patent Application Filings Prediction
The patent applications filings (PAF) are complex to conduct due to its nonlinearity of influenced factors. A novel way about nonlinear regression modeling of PAF with the potential support vector machines (P-SVM) is presented in this study. Grey theory is a truly multidisciplinary and generic theory that deals with systems that are characterized by poor information. However, the application combining grey system theory and P-SVM for PAF prediction is rare. In this study, a grey potential support vector machines with genetic algorithms (GPSVMG) is proposed to forecast PAF. In addition, Grey system is used to add a grey layer before neural input layer and white layer after P-SVM layer. Genetic algorithms (GAs) are used to determine free parameters of P-SVM. The experiments show that the GPSVMG model is outperformed not only single grey model but also SVM with genetic algorithms (SVMG) model and potential support vector machines with genetic algorithms (PSVMG) model. So, PAF prediction based on GPSVMG is of validity and feasibility.
S. Xu and T. Qi, "An Improved P-SVM Model for Patent Application Filings Prediction", Advanced Materials Research, Vols. 255-260, pp. 2141-2144, 2011