Christopher Robinson
Adv. Artif. Intell. Mach. Learn., 3 (1):778-815
Christopher Robinson : NA
DOI: 10.54364/AAIML.2023.1150
Article History: Received on: 23-Jan-23, Accepted on: 25-Feb-23, Published on: 04-Mar-23
Corresponding Author: Christopher Robinson
Email: ckevinr@gmail.com
Citation: Christopher Robinson (2023). Goal Agnostic Learning and Planning without Reward Functions. Adv. Artif. Intell. Mach. Learn., 3 (1 ):778-815
In this paper we present an algorithm, the Goal Agnostic Planner (GAP), which combines elements of Reinforcement Learning (RL) and Markov Decision Processes (MDPs)
into an elegant, effective system for learning to solve sequential problems. The GAP algorithm does not require the design of either an explicit world model or a reward function
to drive policy determination, and is capable of operating on both MDP and RL domain
problems. The construction of the GAP lends itself to several analytic guarantees such as
policy optimality, exponential goal achievement rates, reciprocal learning rates, measurable
robustness to error, and explicit convergence conditions for abstracted states. Empirical results confirm these predictions, demonstrate effectiveness over a wide range of domains, and
show that the GAP algorithm performance is an order of magnitude faster than standard
reinforcement learning and produces plans of equal quality to MDPs, without requiring
design of reward functions.