Development and Comparative Analysis of Machine Learning Models for Urban Traffic Prediction
Year: 2026
Type: article
Abstract: Urban traffic congestion is a major issue in modern cities. Rapid urbanization and increasing vehicle numbers intensify this problem. Accurate traffic prediction is essential for reducing congestion. It also improves road safety and supports intelligent transportation systems. This study develops and compares machine learning models for urban traffic classification. The dataset includes traffic volume indicators and related parameters. Data preprocessing involved exploratory data analysis, feature extraction, and normalization. Three algorithms were implemented: Logistic Regression, Support Vector Machine (SVC), and Random Forest. The experimental results show that Logistic Regression achieved an overall accuracy of 0.87, with weaker performance in the “low” traffic class. The SVC model slightly improved the results, reaching an accuracy of 0.88 and demonstrating better balance across classes. The Random Forest model showed the highest performance, achieving an overall accuracy of 0.99 and weighted F1-score of 0.99, with near-perfect precision and recall for all traffic categories. The findings indicate that ensemble learning methods, particularly Random Forest, provide superior accuracy and reliability for urban traffic prediction tasks. The proposed model can be integrated into real-time traffic management systems to enhance decision-making and improve urban mobility.
Language: English
Subfield: Transportation
Field: Social Sciences
Domain: Social Sciences
Open Access status: bronze
OpenAlex ID: W7142793652