Control What You Can: Intrinsically Motivated Task-Planning Agent
This work addresses the challenge of sample-efficient learning for task-planning agents in robotics, though it appears incremental by combining existing structures with intrinsic motivation.
The paper tackles the problem of enabling an agent to learn how to control its environment efficiently from scratch by optimizing learning progress, resulting in considerably improved performance and smaller sample complexity in synthetic and robotic manipulation tasks.
We present a novel intrinsically motivated agent that learns how to control the environment in the fastest possible manner by optimizing learning progress. It learns what can be controlled, how to allocate time and attention, and the relations between objects using surprise based motivation. The effectiveness of our method is demonstrated in a synthetic as well as a robotic manipulation environment yielding considerably improved performance and smaller sample complexity. In a nutshell, our work combines several task-level planning agent structures (backtracking search on task graph, probabilistic road-maps, allocation of search efforts) with intrinsic motivation to achieve learning from scratch.