Paper Title
Meal Delivery Forecasting Using Machine Learning Models: A Comparative Analysis
Abstract
Day by day supply chain is becoming more and more competitive. It is now necessary for organizations to
accurately predict their customers' behavior to deal with unexpected surges and stillness in the demands of products. With
new approaches to data collection coming up, the volume of data generated is large in number and a variety of nature. So, it
becomes difficult for traditional methods to make forecasts; for the same reason, the idea of using more recent Machine
Learning techniques is being explored and analyzed in this paper. The four models, namely, Linear Regression, Neural
network, Random Forest, and Decision Tree, are applied to forecast demand for a meal delivery company. Three
performance metrics, namely, mean absolute error, mean squared error, and variance score are used to evaluate the
performance of each model. The results showed that the random forest algorithm performed better among all the models
with MAE value of 69.2168, MSE value of 21672.019, and explained variance score value of 0.859.
Keywords - data analytics, machine learning, supply chain, demand forecasting