Research On The Algorithm Of Prediction For The Financing Ability Of Philanthropic Crowdfunding Based On Machine Learning

The rapid development of Internet technology and social media has not only brought about profound changes for the society,but also brought new opportunities for the development of public welfare.With the development of e-payment,the donation for the public welfare is becoming more and more convenient.philanthropic crowdfunding,as a new way for people to donate online,not only brings a new way to setup a project,but also broad the margin of public welfare.Involving in both sponsor and the donator makes public welfare more equal and open,thus promote the development of charity.In such a situation,more and more commonweal organizations turn to the crowdfunding platform,hoping to get more social attention and support through the network market,which brings public welfare into another fierce competing market.At present,the research points of philanthropic crowdfunding focus on its meaning,characteristics and development,few of reserachers look into it through its financing ability.Howerver,the financing ability determines weather the project can be carried out smoothly,and it is also an important basis for crowdfunding platform to optimize the management of philanthropic crowdfunding projects.In view of this,analyzing the main factors which influence the ability of project financing,and then making accurate prediction of project financing ability to promote the rapid development,raising public welfare to improve the project sponsors and investors’ enthusiasm,guaranteeing project the smooth development have important theoretical and practical significance according to the characteristics of different types to raise public welfare projects.This paper focuses on philanthropic crowdfunding,combines with the current development,analyzing the open information of various platform,and research into the prediction of financing ability.Thus,the paper is spitted into 6 chapter.The first chapter is introduction,which put forward the background and the meaning of the research,reviews the domestic and foreign literature,and list the main pointi of the fields of public welfare and crowdfunding.The second chapter is about the concept of philanthropic crowdfunding and the basic theory of it.In this chapter,philanthropic crowdfunding is compared with the traditional public welfare,and the ways to predict its financing ability are introduced.The third chapter is about the construction of the variable of the models,including the condition variables and the aim function,based on the analysis of platform data.The forth chapter is about the preprocessing of data.In this chapter,non-numeric indexes are transformed into numeric ones,the empty datasets are dropped,the abnormal datasets are processed,and the indexes are filtered to fit more to the models by feature engineering.The fifth chapter is to use machine learning models to predict the financing ability of philanthropic crowdfunding,including SVM,RF and GBDT.The last chapter is to conclude the paper,making suggestions for philanthropic crowdfunding,and looking forward to new research.