Overview. This competition aims to design AI/ML methods to solve logistics problems such as variants of the Traveling Salesman Problem (TSP). This year, the problem to solve is a time-dependent orienteering problem with stochastic weights and time windows. It is a problem similar to TSP where nodes need to be visited in a specific order. The goal is to find a route through a network that maximizes some reward. Participants are welcome to join one or two tracks, which each uses a different type of machine learning approaches: (1) online supervised learning/surrogate models and (2) reinforcement learning. More information and how to participate: https://www.tspcompetition.com/
Laurens Bliek (TU Eindhoven, NL), Tom Catshoek (TU Delft, NL), Paulo de Oliveira da Costa (TU Eindhoven, NL), Reza Refaei Afshar (TU Eindhoven, NL), Daniël Vos (TU Delft, NL), Sicco Verwer (TU Delft, NL), Yingqian Zhang (TU Eindhoven, NL).