| The paper started with systematical analyses on relevant articles which targeted the risk warning indicators on commercial bank operation; followed by raising the study purpose of this paper-the research on the early risk warning indictors on Commercial Banking industry based on the favor of SVM. Next, the paper briefly introduced the main types of risk facing the Commercial Banking industry; then analyzed the causes of those risks from two perspectives: external factor-macroeconomic environment, and internal factor-Commercial Banks’own operating systems. Based on the analyses, the paper further built up the structure for commercial banks of key risk warning indictors, which, include the factors that reflect macro-economic environment and the ones that reflect the banks’own financial and operating conditions. In the model-building process, firstly, selected banks were classified according to factor analysis method. Next, the paper projected future data for each group. When using SVM to forecast, after taking into consideration the optimization of several factors, the result of comparison suggested that the accuracy rate on projection is the highest when GA algorithm and PSO algorithm were conducted.The innovations in the paper are as follows:(1) The paper analyzed the internal and external causes of the risks facing Commercial Banks, along with the main types of risks-for instance, capital adequacy risk, credit risk, P&L risk, liquidity risk; followed by constructing a safe and sound risk warning indicating system; (2) When classifying selected banks, the paper implemented factor analysis approach in order to ensure the characteristics of each group would be similar to the real cases; (3) While building up SVM model, the paper compared the merits and shortages among three different algorithms:,GA, and PSO. The paper further tested the GA and PSO on their superiority on optimizing SVM parameters. Overall, the paper provided an effective approach on constructing... |
PDF Full Text Request