Reinforcement learning, compositionally
Author bakirtzisg
9 Stars
Updated Last
1 Year Ago
Started In
June 2022

AlgebraicRL.jl Compositional reinforcement learning.

AlgebraicRL.jl is a Julia library for composing Markov decision processes (MDPs) and their agents compositionally. The library implements algebras for compositional reinforcement learning assisted by Catlab.jl to build a computational interpretation of the categorical semantics that make composition formally precise. The documentation describes the process to compose unit MDPs.

Used By Packages

No packages found.