A Venture Capital-Centric Hybrid Machine Learning Framework for Predicting Startup Success in the Us Ecosystem

Abstract

Venture capitalists facilitate the acquisition of funding for projects, ventures, or causes by aggregating contributions from a substantial pool of investors. This approach enhances the accessibility of financial resources, empowering creators with direct oversight of their funding, devoid of intermediaries such as financial institutions. The adoption of this funding model encourages innovation, bolsters local enterprises, and enables the realization of novel ideas that traditionally struggle to secure financing through conventional methods. By employing an extensive dataset encompassing funding details, feasibility assessments, networking opportunities, and geographical data pertaining to numerous startups, we advance to the stages of analysis and feature engineering, conducting exploratory analysis to discern the various determinants of success. Models were developed using Gradient Boosting, Ada Boost, XGBoost, Support Vector Machine, Random Forest and Light Gradient Boosting Machine. This research work also elucidates significant factors that influence organizational performance and entrepreneurial success, thereby offering stakeholders an innovative perspective on strategic investment management.