Playing Minecraft with Behavioural Cloning
This work addresses sample-efficient agent training for Minecraft, but it is incremental as it applies a standard method to a competition setting.
The authors tackled the MineRL 2019 competition by using behavioral cloning to predict human actions in Minecraft, achieving fifth place, and they analyzed how performance varied with training duration and engineering choices.
MineRL 2019 competition challenged participants to train sample-efficient agents to play Minecraft, by using a dataset of human gameplay and a limit number of steps the environment. We approached this task with behavioural cloning by predicting what actions human players would take, and reached fifth place in the final ranking. Despite being a simple algorithm, we observed the performance of such an approach can vary significantly, based on when the training is stopped. In this paper, we detail our submission to the competition, run further experiments to study how performance varied over training and study how different engineering decisions affected these results.