<![CDATA[Urban transport networks are vital components of modern societies, influencing efficiency and safety. This research explores the potential of traffic data as a crucial information source for forecasting and interpreting traffic problems. Using advanced data processing, statistical analysis, and classification algorithms, the study aims to identify and forecast traffic scenarios. With an interdisciplinary approach integrating computer science, statistics, and transportation engineering, the research emphasizes a holistic perspective on traffic concerns. The study involves outlier detection, label encoding, and cutting-edge technologies like GridSearchCV and ensemble modeling. Inspired by flash flood susceptibility research, machine learning models, particularly LightGBM and CatBoost, are applied to predict traffic situations. DecisionTreeClassifier and CatBoostClassifier emerge as top performers, achieving remarkable accuracies. The evaluation goes beyond accuracy, emphasizing the nuanced understanding of algorithm strengths and limitations for effective urban transportation network management]]>
Copyrights © 2023
