<R implementations of functional principal components analysis(fPCA) to forecast fundraising outcomes in crowdfunding.>
This repository contains implementations of methods from a paper dealing with the problem that popular fundraising platforms don't provide tracking analysis tools for users, which are their demands. It aims to provide a more auccurate way to predict the fundraising outcomes compared with the benchmarks, to serve as a guide in helping the creators and backers to make a better plan for their funding projects.
The implementations details here are all described in the paper: https://www.researchgate.net/publication/322173297_Using_past_contribution_patterns_to_forecast_fundraising_outcomes_in_crowdfunding
Online crowdfunding is regarded as an effective way to raise money, but its biggest limitation is that the creators and backers can not predict whether the fund-raising project will succeed. This article mainly solves this problem. We first explored the nature and heterogeneity of fund-raising dynamics, then compared the successful group with the unsuccessful group, followed by exploratory data analysis. And then a new dynamic model will be shown to predict the results of crowdfunding. This model has made great progress compared with the traditional models.