An evolutionary algorithm-based framework for determining crucial features contributing to the success of a start-up

Startups are forming an important trend in a country's economy. It is vital to know about the features responsible for the startup's success or failure from the analytics point of view. Existing literature focuses upon the predictive behavior and has analyzed the features subjectively. True to our knowledge, none of the existing studies discusses feature selection schemes to identify the essential features responsible for a startup's success. This work discusses a framework based on an evolutionary algorithm to identify the crucial features related to the startup's success. We have used Particle Swarm Optimization for feature selection among a wide range of features of the startup. The proposed framework is evaluated upon a publicly available dataset. The selected features match closely with the subjective results of the popular start-up feature analysis studies. Furthermore, the features selected by the proposed framework obtain an exceptional accuracy of about 92.3 % when trained with popular machine learning classification frameworks.