We study the problem of learning cohesive, fine-grained models of player motion. For instance, agents often choose motion sequences with long-term goals in mind, such as achieving a certain strategic position. Conventional policy learning approaches, such as those based on Markov decision processes, generally fail at learning cohesive long-term behavior in such high-dimensional state spaces, and are only effective when myopic planning leads to the desired behavior. The key difficulty is that such approaches use "shallow" planners that only learn a single state-action policy. We instead propose to learn a hierarchical planner that reasons about both long-term and short-term goals, which we instantiate as a hierarchical deep memory network. We showcase our approach in a case study on modeling basketball player trajectories, and show that it generates significantly more realistic trajectories compared to non-hierarchical baselines as judged by professional sports analysts.
© Copyright 2016 Proceedings of the KDD-16 Workshop on Large-Scale Sports Analytics. Veröffentlicht von Eigenverlag. Alle Rechte vorbehalten.