Research And Implementation Of Consumer Finance Aided Decision System Based On GPS Information

With the rapid development of China's economy and the wide use of Internet technology,financial products have been innovating.A professional personal consumption system has been established in the consumer finance of the Internet,but the risk of issuing loans to customers is becoming higher and higher,whether it is risk control,product delivery,or business expansion,and it requires an auxiliary system for professional analysis of the customers.In view of the decision of the secondary marketing and investment area of consumer finance companies,an auxiliary decision-making system is designed in this paper.The system is based on the mobile GPS data from a large number of loan customers with good loan repayment record,and uses the data mining technology to realize the data analysis of the GPS information,and through the visual analysis,It shows the more active areas of customer activities,focusing on regional heat,combining regional population base,regional government policy guidance and other factors,and making use of decision-making model to provide auxiliary decisionmaking suggestions for the secondary marketing and investment areas of the company.In this paper,we analyze the research status of the related fields at home and abroad,establish the research goal,study the related theory and technology,and combine the requirements analysis of enterprise business and technology,and design the system in detail.The main functions of the system are four parts,including landing verification part,algorithm selection part,visualization part,comment analysis part.The landing verification part mainly manages the login verification module,the functions include login user name,password settings,password modification,and new user registration.The algorithm selection part mainly realizes the management of algorithm modules,including the processing of GPS information data,the selection of algorithm,the selection of parameters,and the integration and conversion of data.The visualization part is mainly about the management of the visual module,and three diagrams are selected to realize the summary analysis of the distribution,and the decision suggestions are displayed in the visual Scatter diagram.The comment analysis part is mainly to realize user's function of analyzing,judging and recommending visual results.The B/S structure is applied in the system development.The Flask framework based on Python is used as the background.The Bootstrap framework is used as the front end,which makes the web implementation more succinct.The data processing is the core part,the algorithm applies DBSCAN clustering algorithms,DBSCAN algorithm uses Haversine formula to make the distance calculation more accurate,choosing Ball-tree's data structure reduces time complexity,the CART decision tree algorithm is used for further judgement.The design of this system basically meets the needs of the enterprise in function and performance,and its operation is more stable.The system can provide reference and basis for the further decision of the consumer finance companies.