- author = "Alina Patelli and John Rego Hamilton and Victoria Lush and Aniko Ekart",
- title = "A {GENTLER} Approach to Urban Traffic Modelling and Prediction",
- booktitle = "2022 IEEE Congress on Evolutionary Computation (CEC)",
- year = "2022",
- editor = "Carlos A. Coello Coello and Sanaz Mostaghim",
- address = "Padua, Italy",
- month = "18-23 " # jul,
- keywords = "genetic algorithms, genetic programming, gplearn, Roads, Computational modeling, Urban areas, Transfer learning, Transportation, Predictive models, Prediction algorithms, Transfer Learning, Exploration-Exploitation Tuning, Intelligent Transportation, Traffic Prediction",
- isbn13 = "978-1-6654-6708-7",
-
DOI = "
doi:10.1109/CEC55065.2022.9870273",
- abstract = "Intelligent Transportation aims to usher in a new and improved version of motorised traffic, one that is streamlined, safe and at the heart of the net-zero agenda. Designing and building the urban infrastructure necessary to turn that vision into a reality relies on complex decision making, which often hinges on estimating the dynamics of future traffic through areas of the road network that are yet to be built. Traffic models capable of yielding such estimations, robustly and reliably, are valuable technological tools that urban planners can use to inform their decisions. To that end, we propose a novel algorithm that employs Genetic Programming and Transfer Learning to produce traffic models which accurately predict vehicle flow through a given junction based on readings collected from surrounding areas. We enhance the algorithm with a randomisation mechanism and run a comprehensive experimental study on a segment of the city of Darmstadts road network, in order to investigate the effects of the exploration-exploitation interplay on the generated models prediction accuracy.",
-
notes = "Also known as
\cite{9870273}
https://gplearn.readthedocs.io/en/stable/reference.html#symbolic-transformer hoist mutation. latin hypercube sampling.",
