Predicting Investment Success on Shark Tank India: A Machine Learning Approach

Abstract

Shark Tank is a critical platform for emerging entrepreneurs, offering not only financial investment but also valuable visibility from renowned investors, or "Sharks." While extensive research has been conducted on Shark Tank in Western markets, there is a notable gap in studies focusing on the Indian context. India has one of the fastest-growing economies in the world, with a rapidly expanding startup ecosystem that is significantly different from Western markets. Additionally, Indian entrepreneurs face unique challenges, such as limited access to funding, regional disparities, and a complex regulatory environment, which make the Shark Tank platform particularly impactful for their success. Understanding the specific factors that influence investment decisions in this context—such as investor preferences, market fit, and demographic considerations—can provide invaluable insights for entrepreneurs looking to secure funding. This study addresses the challenge of predicting whether a pitch will receive an offer by leveraging a diverse dataset with input factors from both U.S. and Indian datasets. It employs a range of regression models, a neural network, and transfer learning techniques to adapt insights from U.S. data to the Indian context, improving the model's predictive accuracy across different markets.  Results indicate that the neural network model achieved the highest predictive accuracy, capturing complex interactions better than simpler models. Remarkably, transfer learning provided reasonable results even after removing city and state features, underscoring its adaptability across diverse markets. These findings highlight the potential of advanced machine learning techniques to improve understanding of investment dynamics in emerging economies like India.