Q-learning versus SVM study for green context-aware multimodal ITS stations

Abstract

Intelligent Transportation Systems (ITS) applications can take big advantage of Context Awareness approaches. Parameters such as user mobility, passengers comfort reaction and pollution emission levels (CO2) can enrich such applications during the decision making phase. Moreover, the expanding in ITS services offers great opportunities for travelers to find the best route to reach their destinations with the lowest or fair costs. It can offer a selecting methodology for optimal route that adapted with some processing parameters like CO2 level, ticket cost, waiting or connection times and the overall traveling time plus the comfortability reaction for each means of transportation) in real time environment using Machine Learning (ML) tools like Q-Learning or SVM: Support Vector Machines. This paper aims at conducting a comparison study for green ITS routes (i.e. the lowest CO2 levels). The study compares between Q-Learning and SVM techniques for identifying different variety of routes between two stops as ranked routes from best to lowest based on some traces gathered from some known transportation traces. Reinforcement Q-Learning is applied to validate the first phase in our approach to recommend the best means and SVM is used to validate the prediction phase about the best route among different routes built based on three means of transportation (metro, train and bus).