An Innovative Fuzzy Logic-Based Machine Learning Algorithm for Supporting Predictive Analytics on Big Transportation Data
Carson K. Leung, Jonathan D. Elias, Shael M. Minuk, A. Roy R. de Jesus, Alfredo Cuzzocrea
Abstract
In the current era of high precision monitoring and big data, many public transit users are still suffering from problems caused by transit delays. To help address this problem, we design and develop an innovative fuzzy logic-based machine learning algorithm for supporting predictive analytics on big transportation data to helps detect and predict the delay of some modes of public transport. To demonstrate the usefulness of our machine learning algorithm as a solution to this problem, we use it on heterogeneous data-namely, transit data and weather data-to predict the expected delay of streetcars (aka trolley cars) in the Canadian city of Toronto. To make accurate prediction, our algorithm takes into account multiple factors such us rain, snow, temperature, time of day, and season. Evaluation results show the effectiveness and usefulness of our fuzzy logic based machine learning algorithm for predictive analytics on big transportation data, which is promising toward development of a predictive intelligent transport system (ITS).