Stochastic control in queueing systems and stochastic networks where specific and stylized model information is used, has a rich history with successes including optimal scheduling and resource allocation in wireless networks and computing systems. In recent times and current applications, model-free approaches such as reinforcement learning, are finding greater applicability owing to increased availability of data of real-world systems and also increased computational capability. To have impact on real-world systems, incorporating model-knowledge and learnings into model-free approaches is necessary, but this is a challenging task. This workshop aims to bring together researchers with expertise in reinforcement learning theory and others with expertise in control in queueing systems and stochastic networks, with a goal to initiate discussions to bridge this challenge.
Please visit HERE ( https://wiopt.online/workshops/track/ReStoq ) for the Technical Program of ReStoq.