Researchers On User’s Behaviors Of Loan Again On Online P2P Lending Platform Under The Ground Of Big Data

Peer to peer lending is a significant component of the new financial system,and is beneficial to solve the problem of financing difficulties of small and medium-sized enterprises,which contributes to the economic development of our country.However,many platforms face with survival crisis because of high operating cost and lack of scientific marketing strategy.They deeply invest on online and offline market promotion but with less effect.In order to solve these problems,peer to peer lending platforms need to analysis and predict users’ behaviors with the help of big data technology,which is conducive to understand customer requirements and make specific marketing strategy.Researchers on user’s behaviors of loan again on online P2 P Lending Platform predict users’ needs of loan again,which will improve the marketing conversion rate and improve its profitability.Aiming at peer to peer lending platform loan again prediction,the following content is studied.Combining the theory of Peer to peer lending and precision marketing,the definition of targeted loan is proposed and the process of targeted lending marketing under the ground of big data have been constructed.Secondly,indicator system of user’s behavior of loan again at peer to peer lending platform have been constructed after analyzing and scouting out users’ data.Indicators screening by elasticated net is applied to make feature selection and select important indexes.And then,Sparse Bayesian Networks is applied to social sciences and results shows it is superior to general Bayesian Networks,which reflect the relationship and important characteristics of user behaviors by analyzing the relationship between the nodes.Next,Xg-boost algorithm is applied to make predictions to whether borrowers could make a second loan at the platform.Experiment result shows Xg-boost algorithm is superior to other algorithms in classification accuracy and time complexity.Finally,suggestions are making for the precision lending marketing of peer to peer lending platform.The main conclusions of this dissertation:(1)Researched on User’s Behaviors of Loan Again on Online P2 P Lending Platform make contribute to make marketing strategy,which is of significant to the strategy making of platform.(2)Elasticated net for indicator screening could decrease redundant variables and improve model performance.(3)The introduction of consumption information and social capital couldimprove the classification accuracy of the models.(4)the prediction performance of Xg-boost algorithm is superior to other algorithms in classification accuracy and time complexity.