A Research On Risk Warning Model Of Mobile Banking Based On BP Neural Network

The developmentof information technology based on mobile internetplatform speeds upthe development process of mobile banking.Because of its popularity and convenienceadvantages, the size of its market and users are constantly expanding. Due to its dual bankingand IT technical characteristics, mobile banking not only faces the risk from conventionalcommercial banks, but also faces the risk based on web information technology and the riskcaused by web-based financial services business. Therefore, compared totraditionalbanks,mobile bankingrisk presents to a moreserious challenge tothebankingregulatory system. Thus, in the development of mobile banking, it’s imperative tounderstand the risks of mobile banking，prevent mobile banking risks and ensure the securityof the country’s financial operation.Based on a comprehensive analysis of mobile banking,14types of risks that affectmobile banking security was identified and the risk hologram of mobile banking was built.After risk primaries, six kinds of principal risks was selected and the early-warning indexsystem of evaluation of mobile banking risk was constructed. Principal component analysismethod was used to analyze450sets of data sample obtained from survey questionnaires, andvalue at risk of virtual warning banks and sample banks was calculated. BP neural networkwas used to test principal component analysis results and the early warning of mobile bankingrisk was conducted, thus mobile banking is divided into two categories of "Security Bank"and "Risk Bank”. Compared with the application of linear expressions, the experimental resultshows that using the BP neural network to forecast mobile banking risk is high accuracy andis feasible.