Ai Driven Sustainable Prediction Model for Start Up Success Using Advanced Machine Learning Techniques

Abstract

India has developed one of the world’s largest start-up ecosystems, with over 1,00,000 registered start-ups and more than 100 unicorns, supported by initiatives such as Startup India and growing venture capital investments. Despite this rapid growth, nearly 85–90% of start-ups fail within the first five years due to issues like poor planning, inadequate funding decisions, and market misalignment, emphasizing the need for intelligent decision-support systems.Traditionally, start-up success is evaluated through manual methods including expert opinions, market surveys, historical case studies, and financial analysis. These approaches are largely subjective, time-consuming, and dependent on human judgment, making them less reliable for handling large and complex datasets or adapting to dynamic market conditions.To address these limitations, the proposed system uses machine learning techniques to predict start-up success or failure based on multiple factors such as funding, milestones, team strength, and market conditions. Using a Kaggle dataset, the data is pre-processed through missing value handling, ADASYN-based class balancing, and feature normalization.Several machine learning models including Random Forest, KNN, SVM, Naïve Bayes, and Logistic Regression are trained and evaluated using performance metrics such as accuracy, precision, recall, F1-score, ROC curve, and confusion matrix. Among them, Random Forest achieves the best performance. Finally, a Flask-based web application is developed to provide user-friendly predictions and investment insights, enabling data-driven decision-making for investors and entrepreneurs