Extracting Action Sequences from Texts Based on Deep Reinforcement Learning.
Wenfeng FengHankz Hankui ZhuoSubbarao KambhampatiPublished in: IJCAI (2018)
Keyphrases
- action sequences
- reinforcement learning
- initial state
- plan recognition
- human actions
- optimal policy
- markov decision processes
- state space
- spatio temporal
- three dimensional
- reasoning about actions
- machine learning
- action recognition
- natural language
- learning algorithm
- low level
- context aware
- activity recognition
- partially observable