Issue |
ITM Web Conf.
Volume 70, 2025
2024 2nd International Conference on Data Science, Advanced Algorithm and Intelligent Computing (DAI 2024)
|
|
---|---|---|
Article Number | 03021 | |
Number of page(s) | 7 | |
Section | Image Processing and Computer Vision | |
DOI | https://doi.org/10.1051/itmconf/20257003021 | |
Published online | 23 January 2025 |
Advanced Network Traffic Prediction Using Deep Learning Techniques: A Comparative Study of SVR, LSTM, GRU, and Bidirectional LSTM Models
College of Arts and Sciences, New York University, New York, United States
Corresponding author: yw6054@nyu.edu
Accurate prediction of network traffic patterns is essential for optimizing network resource allocation, managing congestion, and strengthening cybersecurity. This study examines the effectiveness of four machine learning models—Support Vector Regression (SVR), Long Short-Term Memory (LSTM), Gated Recurrent Units (GRU), and Bidirectional Long Short-Term Memory (Bi-LSTM)—in forecasting traffic patterns using both web-based and real-world datasets. The models are evaluated based on their generalization accuracy, as measured by Mean Absolute Percentage Error (MAPE), computational efficiency, and their ability to capture underlying traffic dynamics. Results indicate that GRU surpasses SVR and LSTM in terms of prediction accuracy and computational speed, while Bidirectional LSTM demonstrates superiority in capturing long-term dependencies across extended periods. These findings underscore the significant potential of deep learning models, particularly GRU and Bidirectional LSTM, in improving the precision and reliability of network traffic predictions. The study offers insights into the strengths and limitations of each model, contributing to the ongoing development of more robust and efficient network traffic forecasting methods.
© 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.