A Recommendation Model for Adding Locations for a Food Delivery Service

This paper presents our approach to making recommendations for adding locations to a restaurant food delivery service. The aim of the paper is to suggest restaurants to connect to the service in terms of improving its financial efficiency. To achieve this goal, we studied existing options for implementing similar tasks. In our case, to build recommendations, we trained a classification model that determines whether a restaurant is financially efficient to add, based on historically added restaurants to the service. The training dataset included information about the category and segment of a restaurants, information about the area where restaurants are located, and information from popular geo-services about these restaurants. Using the trained model, we created a system that recommends restaurants to add to the service based on the current day data.