JJEM: Special Edition 2 (December 2024)
JJEM: Special Edition 2 (December 2024)
2024-12-08
Startup Success Prediction Using Machine Learning Algorithms
Dr. Harish B. G, Achuth Kumar R Ranjith Kumar C K
The importance of startups for a dynamic, innovative and competitive economy has already been acknowledged in the scientific and business literature.. The startup ecosystem is characterized by high levels of uncertainty and volatility, which makes the process of evaluating company success through information analysis and interpretation laborious and computationally demanding. This prediction dilemma highlights the necessity for a quantitative methodology that should allow for an unbiased, fact-based method of predicting startup success. The information utilized in this analysis was obtained from crunchbase.com, an online investment platform. The oversampling technique, ADASYN, has been used to pre-process the data for sampling bias and imbalance. Four distinct models are employed in order to forecast the success of a startup. The best models chosen are the ensemble approaches, random forest, and extreme gradient boosting, with test set prediction accuracy of 94.1% and 94.5%, respectively, using goodness-of-fit metrics that are applicable to each model situation.
Decision tree, Random forest, K-Nearest Neighbors (KNN)