Demand Optimization Forecasting Using Machine Learning for E-Commerce Platform Recommendation

Abstract

The paper presents an intelligent platform recommendation system based on machine learning to ensure optimal demand management through appropriate platform selection. The selection of appropriate marketplaces is needed to enhance product visibility, sales, and decision-making. The data collection process is performed via web scraping using Selenium, while selling price, MRP, discount, rating, rating count, and subcategory are important features used to help determine the optimal platform for the product's sale. The collected data is then preprocessed involving feature engineering and then classified using Random Forest and XGBoost classifiers. Performance evaluation is done using stratified cross-validation. The experimental results show that the XGBoost classifier outperforms the Random Forest classifier in terms of accuracy and consistency. Moreover, the proposed framework has been integrated into a web application built using Flask.