Injecting Prior Knowledge for Transfer Learning into Reinforcement Learning Algorithms using Logic Tensor Networks.
Samy BadreddineMichael SprangerPublished in: CoRR (2019)
Keyphrases
- transfer learning
- reinforcement learning algorithms
- reinforcement learning
- prior knowledge
- tabula rasa
- labeled data
- learning tasks
- learning algorithm
- knowledge transfer
- function approximation
- model free
- state space
- temporal difference
- cross domain
- unlabeled data
- reinforcement learning methods
- higher order
- machine learning
- markov decision processes
- machine learning algorithms
- semi supervised learning
- training data
- optimal policy
- transfer knowledge
- reward function
- network structure
- text mining
- target domain
- collaborative filtering
- data sets
- active learning
- supervised learning
- multi task
- text classification
- text categorization
- dynamic programming
- random walk
- pairwise
- function approximators
- dynamic environments
- monte carlo