Teaching AI Agents Ethical Values Using Reinforcement Learning and Policy Orchestration.
Ritesh NoothigattuDjallel BouneffoufNicholas MatteiRachita ChandraPiyush MadanKush R. VarshneyMurray CampbellMoninder SinghFrancesca RossiPublished in: IJCAI (2019)
Keyphrases
- reinforcement learning
- action selection
- agent receives
- multi agent
- optimal policy
- sequential decision making
- learning agents
- agent learns
- policy search
- learning process
- multi agent systems
- reward function
- multiple agents
- learning agent
- intelligent agents
- artificial intelligence
- function approximation
- action space
- intelligent behavior
- multiagent systems
- reinforcement learning algorithms
- temporal difference
- single agent
- multi agent reinforcement learning
- cooperative
- software agents
- function approximators
- markov decision processes
- robocup soccer
- state space
- learning capabilities
- expert systems
- machine learning
- intelligent systems
- actor critic
- multiagent learning
- rl algorithms
- evolutionary learning
- dynamic environments
- decision making
- reinforcement learning agents
- ai researchers
- autonomous agents
- model free
- technology enhanced learning
- policy iteration
- control policy
- online learning
- resource allocation
- expected reward
- state and action spaces
- markov decision process
- reinforcement learning problems
- multi agent environments
- web services
- policy evaluation
- policy gradient
- learning processes
- complex environments
- inverse reinforcement learning
- dynamic programming
- partially observable
- multiagent reinforcement learning
- partially observable markov decision processes
- partially observable environments
- reward signal
- markov decision problems