Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 02014 | |
Number of page(s) | 12 | |
Section | Machine Learning in Healthcare and Finance | |
DOI | https://doi.org/10.1051/itmconf/20257002014 | |
Published online | 23 January 2025 |
Enhancing Predictive Models in E-Commerce: A Comparative Study Using XGBoost Across Diverse Scenarios
Beijing Royal School No.11 Wangfu Street, North Qijia Town, Changping District, Beijing, 102207, China
Corresponding author: YFSLZYS@163.com
With the growth of the internet, online shopping has become increasingly popular. However, sudden demand spikes during holidays or special events can disrupt market equilibrium, causing stock shortages and logistical challenges. To address these sudden surges in demand, this study utilizes existing online sales data, transforming it into actionable insights. Our strategy involves continuously feeding historical data into selected models to predict future sales volumes. By identifying patterns in the data, we aim to make the predictions more tangible and assess the validity of our approach through a statistical linear regression model. We employed three different models—Gradient Boosting Decision Tree (GBDT), Light Gradient Boosting MachineLight (GBM), and Extreme Gradient Boosting (XGBoost)—to determine which one is the most efficient. After comparing the performance of these models, the results indicate that XGBoost is the optimal choice. The findings suggest that accurate sales predictions enable e-commerce platforms to increase inventory during peak periods, maximize the utility of goods, ensure customer satisfaction, and stimulate transaction activity. This study underscores the importance of accurately forecasting sales and revenue in e-commerce, helping platforms to stay ahead of demand, optimize resource allocation, and maintain market competitiveess.
© The Authors, published by EDP Sciences, 2025
This is an Open Access article distributed under the terms of the Creative Commons Attribution License 4.0, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.