Risk Assessment of Processing Cross Border Ecommerce Using Improved Radial Basis Function Kernel-Support Vector Machine in Data Mining

Data mining is a process of determining trends, patterns, and insights to extract significant data for decision-making which helps in better understanding and data utilization. Internet technology has been established and spreads primarily by utilizing Internet for financial management become a new kind of financial technology. Internet finance’s profits and convenience, severe financial risks like loans and chaotic reputation management have appeared. To increase the cross-border e-commerce financial ability of enterprises to respond to over-financialization, the Improved Radial Basis Function kernel-Support Vector Machine (IRBF kernel-SVM) is proposed based on data mining in this research. Initially, data is obtained from a processing of cross border ecommerce enterprise with Chinese banks. The pre-processing using deleting and filling methods are utilized for missing values and outlier processing which improves the model performance. Then, the Chi-square is employed to select the relevant features and IRBF kernel-SVM is performed for risk assessment of processing cross border e-commerce. The proposed IRBF kernel-SVM achieves a better accuracy of 0.97 compared to the existing techniques like Decision Tree (DT), Random Forest (RF), and SVM respectively.