Traffic signal control is an important problem in urban mobility with a significant potential for economic and environmental impact. While there is a growing interest in Reinforcement Learning (RL) for traffic signal control, the work so far has focussed on learning through simulations which could lead to inaccuracies due to simplifying assumptions. Instead, real experience data on traffic is available and could be exploited at minimal costs. Recent progress in offline or batch RL has enabled just that. Model-based offline RL methods, in particular, have been shown to generalize from the experience data much better than others.

We build a model-based learning framework that infers a Markov Decision Process (MDP) from a dataset collected using a cyclic traffic signal control policy that is both commonplace and easy to gather. The MDP is built with pessimistic costs to manage outof-distribution scenarios using an adaptive shaping of rewards which is shown to provide better regularization compared to the prior related work in addition to being PAC-optimal. Our model is evaluated on a complex signalised roundabout and a large multiintersection environment, demonstrating that highly performant traffic control policies can be built in a data-efficient manner