A novel methodology to predict urban traffic congestion with ensemble learning

A Comparative Study of Ensemble Models for Predicting Road Traffic Congestion

Increased road traffic congestion is due to different factors, such as population and economic growth, in different cities globally. On the other hand, many households afford personal vehicles, contributing to the high volume of cars. The primary purpose of this study is to perform a comparative analysis of ensemble methods using road traffic congestion data. Ensemble methods are capable of enhancing the performance of weak classifiers. The comparative analysis was conducted using a real-world dataset and bagging, boosting, stacking and random forest ensemble models to compare the predictive performance of the methods. The ensemble prediction models are developed to predict road traffic congestion. The models are evaluated using the following performance metrics: accuracy, precision, recall, f1-score, and the misclassification cost viewed as a penalty for errors incurred during the classification process. The combination of AdaBoost with decision trees exhibited the best performance in terms of all performance metrics. Additionally, the results showed that the variables that included travel time, traffic volume, and average speed helped predict vehicle traffic flow on the roads. Thus, the model was developed to benefit transport planners, researchers, and transport stakeholders to allocate resources accordingly. Furthermore, adopting this model would benefit commuters and businesses in tandem with other interventions proffered by the transport authorities.

Combining predictive base models using deep ensemble learning

Since information science and communication technologies had improved significantly, data volumes had expanded. As a result of that situation, advanced pre-processing and analysis of collected data became a crucial topic for extracting meaningful patterns hidden in the data. Therefore, traditional machine learning algorithms generally fail to gather satisfactory results when analyzing complex data. The main reason of this situation is the difficulty of capturing multiple characteristics of the high dimensional data. Within this scope, ensemble learning enables the integration of diversified single models to produce weak predictive results. The final combination is generally achieved by various voting schemes. On the other hand, if a large amount of single models are utilized, voting mechanism cannot be able to combine these results. At this point, Deep Learning (DL) provides the combination of the ensemble results in a considerable time. Apart from previous studies, we determine various predictive models in order to forecast the outcome of two different case studies. Consequently, data cleaning and feature selection are conducted in advance and three predictive models are defined to be combined. DL based integration is applied substituted for voting mechanism. The weak predictive results are fused based on Recurrent Neural Network (RNN) and Long Short Term Memory (LSTM) using different parameters and datasets and best predictors are extracted. After that, different experimental combinations are evaluated for gathering better prediction results. For comparison, grouped individual results (clusters) with proper parameters are compared with DL based ensemble results.