This was done as part of Engineering Final year project.

Since there are not many publicly availble free data sets for this task, we have used DonorsChoose.org dataset uploaded on Kaggle.

DonorsChoose.org receives hundreds of thousands of project proposals each year for classroom projects curated by the teachers and coordinator which are in need of funding. As the organization’s reach scaled, a large number of volunteers are required to manually screen each submission before it's approved. The problem we have to solve is How to scale current manual processes and resources to screen 500,000 projects so that they can be posted as quickly and as efficiently as possible.

We aim to use state of the art machine learning techniques in order to train our model to automatically predict the acceptance or rejection of the project proposals thereby reducing the amount of human work employed. Apart from that we also aim to perform various data analysis tasks to find out relevant insights and gain good business outcomes from it.

The principles used in this project here can be similarly be applied to other crowdfunding platforms and changed according to their domain.