Predicting Supermarket Sales Data Using Machine Learning Algorithms

Abstract

The prediction of data in domains such as medical, marketing, sales, weather or stocks is having huge demand now a days. Prediction of sales is plays a vital role in todays world. In order to predict sales, an implementation of several algorithms is done in this work. The data for this work comprises of weekly sales data from different sectors in Supermarket, which exists all over the United States of America. The models implemented here for the prediction are Gradient Boosting, Random Forest and Extremely Randomized Trees Classifiers. To discover the best algorithm and the additional parameter values at which the best outcomes are obtained, a comparative assessment of the three algorithms is used.