A machine learning approach to suggest ideal geographical location for new restaurant establishment

Abstract

Restaurant business is a prospective and profitable business nowadays. However, ensuring quality food, good stuff, inner-environment etc. is a big concern and most importantly before facing all these, the trickiest part is to choose a perfect place where a restaurant business will flourish. Without doing a perfect research on this area, setting up a restaurant may lead to an immediate downfall. In recent time, for choosing a preferred restaurant location and calculating the estimated risk, people are now hiring professionals to do ground check and here the data scientists are coming into play as a bigshot. This research is focused on suggesting a suitable place for setting up a restaurant business based on the existing data from Yelp where 75 features have been extracted for supervised machine learning. Our model will also calculate the expected rating that a restaurant will get depending on the features the restaurant possesses. Several machine learning algorithms (Support Vector Machine, Decision Tree, Logistic Regression and Decision Tree with presort) have been used and juxtaposed to nurture out the suitable one. As yelp's review is authentic and it is maintained regularly, we have considered the rating of a business as the point of suggestion. We have also looked at the comparative analysis of these algorithms and searched for an algorithm that gives us the best result.