Authors : Poornima R. D.; Sumukh M. S.; Varshini H. Gowda; Spoorthi K.; Srikanth P. V.

Volume/Issue : Volume 10 - 2025, Issue 11 - November

Google Scholar : https://tinyurl.com/y6yxxeaw

Scribd : https://tinyurl.com/33ywe5tz

DOI : https://doi.org/10.38124/ijisrt/25nov1062

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Note : Google Scholar may take 30 to 40 days to display the article.

Abstract : Investing in new startups is a high-risk endeavor often reliant on 'gut feeling'—a method that isn't always accurate. This paper presents a system to support investor decisions using data. We built a system that collects key data about a startup—like its funding, industry, and team size—and uses an AI model to predict if it's likely to succeed (be acquired) or fail (close). For this "Prediction Engine," we developed a Stacked Ensemble (XGBoost, LightGBM, RandomForest). We picked this architecture because it provides stable, high-performance predictions. In our testing, this model proved to be very effective, achieving a 79.5% accuracy rate and, more critically, a 92.5% Recall rate, minimizing the high cost of missing a successful startup. The primary contribution of this work is not only the development of a high- recall predictive pipeline but also its commitment to transparency. We move beyond the 'black box' paradigm by implementing SHAP (SHapley Additive exPlanations) to provide full model interpretability. This analysis reveals the specific, non-linear drivers of success, such as 'funding momentum' and 'milestone velocity.' The entire tool is a full-stack website built with React for the frontend, Flask (Python) for the backend, and MongoDB for the database. Our main goal was to take all that complicated data and make it simple, clear, and easy to understand, so people can make decisions based on facts, not just hunches.

Keywords : Startup Success Prediction, Machine Learning, Stacked Ensemble, Interpretable AI, SHAP, Venture Capital, Decision Support System, Full-Stack Development.

References :

1. A. Krishna, A. Agrawal, and A. Choudhary, “Predicting the Outcome of Startups: Less Failure, More Success,” Proc. IEEE ICDM Workshops, 2016.
2. A. K. Misra, D. S. Jat, and D. K. Mishra, “Startup Success and Failure Prediction Using k-Means Clustering and Artificial Neural Network,” Proc. IEEE ETNCC, 2023.
3. C. Ünal and I. Ceasu, “A Machine Learning Approach Towards Startup Success Prediction,” IRTG 1792 Discussion Paper No. 2019-022, Humboldt University, Berlin, 2019.
4. M. R. Bidgoli, I. R. Vanani, and M. Goodarzi, “Predicting the Success of Startups Using a Machine Learning Approach,” J. Innovation and Entrepreneurship, vol. 13, no. 80, 2024.
5. P. X. McCarthy, et al., “The Impact of Founder Personalities on Startup Success,” Nature Scientific Reports, vol. 13, art. 17200, 2023.
6. H. Baskoro, H. Prabowo, Meyliana, and F. L. Gaol, “Predicting Startup Success: A Literature Review,” BINUS University Press, 2021.
7. E. Skawińska and R. I. Zalewski, “Success Factors of Startups in the EU—A Comparative Study,” Sustainability, vol. 12, no. 19, p. 8200, 2020.
8. S. Ahluwalia and S. Kassicieh, “Effect of Financial Clusters on Startup Mergers and Acquisitions,” Int. J. Financial Studies, vol. 10, no. 1, 2022.
9. R. Allu and V. N. R. Padmanabhuni, “A Machine Learning Approach for Predicting Startup Success Using Social Media Data,” Int. J. Innovative Technology and Exploring Engineering (IJITEE), vol. 9, no. 5, 2020.
10. V. Ramakrishna and N. Rao, “SME Success Prediction Using Hybrid ML Models,” IJITEE, vol. 9, no. 5, 2020.
11. H. Gadam, R. K. Pal, and T. A. Desai, “Startup Success Prediction Using GRU-SAM: A Big Data-Driven Financial Modeling Approach with LLM-Enhanced Insights,” Proc. IEEE ICDSIS, 2025.
12. Y. Lisanti, D. Luhukay, and V. Mariani, “IT Service and Risk Management Implementation for Online Startup SMEs,” Proc. IEEE ICIMTech, 2017.
13. C. Zhang, H. Zhang, and X. Hu, “A Contrastive Study of Machine Learning on Funding Evaluation Prediction,” IEEE Access, vol. 7, pp. 1–9, 2019.