Abstract

Urban traffic congestion imposes significant economic, environmental, and social costs on rapidly growing cities worldwide. This research investigates how predictive analytics and machine  learning can be leveraged to classify and forecast traffic congestion severity in real time,  enabling data-driven decision-making for transportation planning, signal optimization, and  congestion management. A real-world traffic monitoring dataset comprising 5,952 observations collected over two months via  computer vision sensors at an urban intersection was analysed under the CRISP-DM frame- work. The  dataset records counts of four vehicle classes including cars, bikes, buses, and trucks at  15-minute intervals, alongside temporal variables such as time of day, date, and day of week.  Analysis was performed using Python 3.11 with scikit-learn, scipy, pandas, and matplotlib, and  cross-validated using IBM SPSS Statistics 29. Four supervised machine learning classification algorithms were trained and evaluated on a  stratified 70/30 train-test split. XGBoost (Gradient Boosting Classifier) achieved the highest  performance with an accuracy of 99.89%, weighted F1-score of 0.9989, precision of 0.9989, and  recall of 0.9989 on the held-out test set of 1,786 observations, substantially outperforming Linear  SVM (90.76% accuracy) and Logistic Regression (88.69%). Random Forest achieved 99.66% accuracy as  the second-best model. Statistical analysis confirms highly significant associations between traffic situation and hour of  day (Chi-Square = 2,413.211, df = 15, p < .001), PM peak period (Chi-Square = 1,138.848, df = 3, p  < .001), and all vehicle count variables (Kruskal-Wallis H values ranging from 98.3 to 3,242.9, all  p < .001). Exploratory analysis reveals that the PM Peak (16:00–18:00)

Publication Date

4-2026

Document Type

Thesis

Student Type

Graduate

Degree Name

Professional Studies (MS)

Department, Program, or Center

Graduate Programs & Research

Advisor

Parthasarathi Gopal

Campus

RIT Dubai

Download

Share

COinS