Enforcing ethical goals over reinforcement-learning policies.
Emery A. NeufeldEzio BartocciAgata CiabattoniGuido GovernatoriPublished in: Ethics Inf. Technol. (2022)
Keyphrases
- reinforcement learning
- optimal policy
- policy search
- markov decision process
- control policies
- state space
- reward function
- markov decision problems
- function approximation
- partially observable markov decision processes
- fitted q iteration
- reinforcement learning agents
- markov decision processes
- hierarchical reinforcement learning
- finite state
- model free
- reinforcement learning algorithms
- transfer learning
- total reward
- decision problems
- learning process
- cooperative multi agent systems
- multiagent reinforcement learning
- reinforcement learning methods
- control policy
- dynamic programming
- action space
- learning problems
- approximate policy iteration
- policy gradient methods
- information systems
- machine learning
- management policies
- temporal difference learning
- state and action spaces