Forecasting Donations Through Machine Learning: Exploring Donation-Centric Crowdfunding in Indian Scenario

In this study we aim to forecast monetary contributions using three machine learning methods based on geographical location, state, required donation amount and number of supporters. Data from Milaap portal between Jan 2022 to Dec 2022 with 615 participants and seven attributes were preprocessed and analyzed using data visualization and machine learning techniques. Findings indicate state-based and gender-related variations in funding distribution. Certain Indian states have fundraising advantages while males benefit more from donation-based crowdfunding. Fundraisers typically secure 40–50% of requested donations. K Nearest Neighbor and Random Forest Regressor were effective for donation prediction. This study addresses a gap in the literature by offering a predictive model for donation-based crowdfunding aiding the estimation of donations based on demographics and geography. However, findings are limited to data obtained from Milaap portal warranting caution when generalizing to other platforms. This extends existing Indian crowdfunding research contributing a predictive approach for donation estimation.