Our paper explores factors that influence financing proportions for crowdfunding projects by modifying various regression models. Our findings demonstrate that offering various types of rewards, such as interview opportunities and product incentives, can increase the likelihood of achieving higher financing proportions. Furthermore, we introduce two tree-based regression models, the Hybrid Tree-based Linear Regression (HTLR) and the Hybrid Forest-based Linear Regression (HFLR), which outperform the traditional linear regression model in both explanatory and predictive power. We attribute this to the inclusion of nonlinear terms and the ability to handle outliers. Our results show that the HFLR model is superior to the HTLR model when an appropriate loop termination condition is applied, due to its ability to combine results from multiple decision trees and account for joint effects of factors.
Dr Hui Li: Dr Li is a Reader in Applied Statistics and Econometrics at the School of Mathematics at the University of Birmingham. She earned her PhD in Economics and Econometrics from the University of New Mexico in the USA in 2003. Before joining the University of Birmingham, she held positions as a Senior Statistician at the New Mexico State Government and an Associate Professor at Eastern Illinois University in the USA.
Dr. Li's research focuses on statistical modelling and has made significant contributions to the fields of health care and medical sciences, environmetrics, and finance. Her research in health care and clinical studies involves developing statistical models to analyse clinical trials and longitudinal studies. Her research in finance focuses on time-series modelling of financial market volatility, and her work in environmetrics includes cost and benefit analysis in policy evaluations.