Attention-Based Response Generation Using Parallel Double Q-Learning for Dialog Policy Decision in a Conversational System.
Ming-Hsiang SuChung-Hsien WuLiang-Yu ChenPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2020)
Keyphrases
- optimal policy
- conversational agent
- decision problems
- action selection
- decision process
- decision processes
- conversational agents
- natural language
- reinforcement learning
- decision making
- decision makers
- state space
- cooperative
- policy iteration
- function approximation
- multi agent
- visual attention
- decision model
- parallel implementation
- state action
- mixed initiative
- learning algorithm
- parallel processing
- multi modal
- markov decision processes
- infinite horizon
- machine learning
- dynamic programming
- stochastic approximation
- decision rules
- markov decision process
- reward function
- shared memory
- utility function
- action space
- reinforcement learning algorithms
- temporal difference learning
- potential field
- decision theory
- continuous state spaces
- visual stimuli
- dialog systems