World Value Functions: Knowledge Representation for Multitask Reinforcement Learning.
Geraud Nangue TasseSteven D. JamesBenjamin RosmanPublished in: CoRR (2022)
Keyphrases
- multi task
- reinforcement learning
- knowledge representation
- transfer learning
- multi task learning
- multitask learning
- learning problems
- learning tasks
- machine learning
- multi class
- knowledge base
- artificial intelligence
- multiple tasks
- multiclass support vector machines
- expert systems
- learning process
- natural language
- markov random field
- supervised learning
- gaussian processes
- binary classification problems
- learning algorithm
- data sets