Analysis and Classification of Restaurants Based on Rating with XGBoost Model

Abstract

The restaurant business is one of the most competitive and the need for restaurants is growing daily. Bangalore is a foodie's paradise, boasting cuisines from all over the world. Hence, this research work focuses the classification of restaurants on the basis of rate with XGBoost model. The Exploratory Data Analysis(EDA) with different graphs provides an analysis of the data before classification. Prior to EDA the data sets are cleaned with various steps to increase the accuracy of the visualization and classification. The research was performed using Zomato data set for restaurants in a specific locality(Bangalore). Data Visualization techniques helped to analyze food culture, trends and patterns. This research describes a model for understanding the elements that influence restaurant ratings. Predictive analytics and machine learning together with a variety of tools and methodologies, help in predicting restaurant ratings. The model in this research is developed using multiple supervised techniques, with the most efficient algorithm being evaluated. The classification XGBoost model provided an accuracy of 98.07 %. The outcome of the work assists new restaurants in selecting on their menu, cuisine, theme, prices, geographic location, and so on, consequently enhancing business. The rate and location details helps people to select the restaurants for the dining.