Reinforcement Learning of Two-Issue Negotiation Dialogue Policies.
Kallirroi GeorgilaPublished in: SIGDIAL Conference (2013)
Keyphrases
- reinforcement learning
- optimal policy
- multi agent
- policy search
- resolving conflicts
- reward function
- function approximation
- cooperative
- control policies
- decision making
- state space
- conflict resolution
- fitted q iteration
- markov decision process
- learning capabilities
- learning process
- model free
- multi agent systems
- reinforcement learning algorithms
- access control policies
- multiagent reinforcement learning
- markov decision problems
- control policy
- continuous state
- dialogue management
- reinforcement learning agents
- markov decision processes
- learning algorithm
- action selection
- mixed initiative
- conversational agent
- policy gradient methods
- machine learning
- automated negotiation
- software agents
- dynamic programming
- partially observable markov decision processes
- access control
- dialogue system
- temporal difference
- long run
- finite state