Develop a classifier to aid the evaluation of startup applications

Abstract

Power of Ideas is the largest entrepreneurship program in India which identifies startups with a strong potential and provides support for their growth. In this project, we focus on developing a classification model that can predict the strength of an application on certain key parameters, that will aid the manual evaluators. We have used cluster extraction to identify keywords which form the core an application. These keywords are later used in developing a multinomial logistic regression model to evaluate an application. On a parallel note, we developed another classifier based on Naïve Bayes algorithm. Through the results of these models, we have identified a few key areas of focus for evaluators as well as applicants to strengthen their application.