Optimizing Sales Prediction in Online Grocery Shopping using Product Embeddings

Abstract

The focus of this research is to delve into the world of product embeddings and their potential to improve prediction accuracy. The study included a thorough review of existing literature to highlight the generation of embeddings and their utility in downstream tasks. This review uncovered the significance of embeddings in enhancing prediction accuracy, showing promise for their application in the realm of online grocery shopping. By employing embedding techniques, we aim to examine how they refine sales predictions, ultimately contributing to improved inventory management and optimized supply chain processes in the e-commerce domain. Future work will encompass data acquisition and a comprehensive process involving embedding generation, model training, and hyperparameter tuning to validate these findings.