Multi-task curriculum learning in a complex, visual, hard-exploration domain: Minecraft

Citations Over Time

Abstract

An important challenge in reinforcement learning is training agents that can solve a wide variety of tasks. If tasks depend on each other (e.g. needing to learn to walk before learning to run), curriculum learning can speed up learning by focusing on the next best task to learn. We explore curriculum learning in a complex, visual domain with many hard exploration challenges: Minecraft. We find that learning progress (defined as a change in success probability of a task) is a reliable measure of learnability for automatically constructing an effective curriculum. We introduce a learning-progress based curriculum and test it on a complex reinforcement learning problem (called "Simon Says") where an agent is instructed to obtain a desired goal item. Many of the required skills depend on each other. Experiments demonstrate that: (1) a within-episode exploration bonus for obtaining new items improves performance, (2) dynamically adjusting this bonus across training such that it only applies to items the agent cannot reliably obtain yet further increases performance, (3) the learning-progress based curriculum elegantly follows the learning curve of the agent, and (4) when the learning-progress based curriculum is combined with the dynamic exploration bonus it learns much more efficiently and obtains far higher performance than uniform baselines. These results suggest that combining intra-episode and across-training exploration bonuses with learning progress creates a promising method for automated curriculum generation, which may substantially increase our ability to train more capable, generally intelligent agents.

Related Papers

• The Design of Curriculum Structure in the Synthesis of School Curriculum(2008)
• Postgraduates' Curriculum-selecting Intention and System Reform(2009)
• The Teachers' Curriculum Identity: An Issue Needs Further Study(2011)
• On the Paradigm Change of Physical Education Curriculum: the Analysis of Three Typical PE Curriculum Models(2004)
• → Comparative analysis of the 2019 revised Nuri curriculum and the Japanese kindergarten curriculum(2023)