PADDLE: Logic Program Guided Policy Reuse in Deep Reinforcement Learning.
Hao ZhangTianpei YangYan ZhengJianye HaoMatthew E. TaylorPublished in: AAMAS (2024)
Keyphrases
- logic programs
- reinforcement learning
- optimal policy
- logic programming
- answer sets
- stable models
- answer set programming
- fixpoint
- normal logic programs
- background knowledge
- logic program updates
- state space
- computational properties
- inside outside algorithm
- markov decision processes
- general logic programs
- inductive logic programming
- prolog programs
- loop formulas
- learning algorithm
- stable model semantics
- dynamic programming
- extended logic programs
- machine learning
- existentially quantified