Interactive Learning of Environment Dynamics for Sequential Tasks
This addresses the challenge of efficient robot learning for sequential tasks by leveraging human expertise, though it is incremental as it builds on existing task inference work.
The paper tackles the problem of robots learning environment dynamics from human teachers by proposing Behavior Aware Modeling (BAM), which incorporates teacher knowledge into transition dynamics models, and shows it outperforms methods that ignore this knowledge in simulations and real human experiments.
In order for robots and other artificial agents to efficiently learn to perform useful tasks defined by an end user, they must understand not only the goals of those tasks, but also the structure and dynamics of that user's environment. While existing work has looked at how the goals of a task can be inferred from a human teacher, the agent is often left to learn about the environment on its own. To address this limitation, we develop an algorithm, Behavior Aware Modeling (BAM), which incorporates a teacher's knowledge into a model of the transition dynamics of an agent's environment. We evaluate BAM both in simulation and with real human teachers, learning from a combination of task demonstrations and evaluative feedback, and show that it can outperform approaches which do not explicitly consider this source of dynamics knowledge.