همایش بین المللی هوش مصنوعی

صفحه اصلی / اولین همایش بین المللی هوش مصنوعی

Empowering Decision-Making in Venture Investments: A Systematic Review of Machine Learning Applications for Predicting Startup Success

نویسندگان :

Seyed Mohammad Javad Toghraee¹ Hadi Nilforoushan² Nafiseh Sanaee³

1- Faculty of Management, Shahid Beheshti University, Tehran, Iran 2- Department of Science and Technology Policy, Institute of Fundamental Studies of Science and Technology, Shahid Beheshti University, Tehran, Iran 3- Faculty of Management, Shahid Beheshti University, Tehran, Iran

کلمات کلیدی :

Startup Success Prediction،Data-Driven Decision-Making،Machine Learning in Venture Capital،Investment Risk Assessment،Machine Learning

چکیده :

Startup success is inherently unpredictable, with most ventures failing within their early years. This systematic review, following PRISMA guidelines, synthesizes findings from 23 peer-reviewed studies on machine learning (ML) applications in predicting startup success. Key ML techniques, including Random Forest, Gradient Boosting, and hybrid models, demonstrated high accuracy (up to 94.3%) across diverse datasets like Crunchbase and Kaggle. Critical success factors identified include funding patterns, team composition, market adaptability, and social media engagement. Relational approaches, such as graph embeddings, underscored the importance of proximity to investors and industry networks. However, reliance on incomplete public datasets and limited integration of qualitative factors remain challenges. This review provides actionable insights for investors, entrepreneurs, and policymakers, highlighting ML's transformative potential in fostering data-driven decision-making. Future research should focus on diversifying datasets, improving explain ability, and integrating qualitative factors to address existing gaps.