A Theory of Abstraction in Reinforcement Learning
This work addresses the challenge of efficient decision-making for reinforcement learning agents by providing a foundational theory and methods for abstraction, though it appears incremental as it builds on existing concepts.
The dissertation tackles the problem of abstraction in reinforcement learning by proposing a theory with three desiderata for abstraction functions and developing new algorithms to learn them, aiming to reduce the complexity of effective reinforcement learning.
Reinforcement learning defines the problem facing agents that learn to make good decisions through action and observation alone. To be effective problem solvers, such agents must efficiently explore vast worlds, assign credit from delayed feedback, and generalize to new experiences, all while making use of limited data, computational resources, and perceptual bandwidth. Abstraction is essential to all of these endeavors. Through abstraction, agents can form concise models of their environment that support the many practices required of a rational, adaptive decision maker. In this dissertation, I present a theory of abstraction in reinforcement learning. I first offer three desiderata for functions that carry out the process of abstraction: they should 1) preserve representation of near-optimal behavior, 2) be learned and constructed efficiently, and 3) lower planning or learning time. I then present a suite of new algorithms and analysis that clarify how agents can learn to abstract according to these desiderata. Collectively, these results provide a partial path toward the discovery and use of abstraction that minimizes the complexity of effective reinforcement learning.