Training and Evaluation of Deep Policies using Reinforcement Learning and Generative Models.
Ali GhadirzadehPetra PoklukarKarol ArndtChelsea FinnVille KyrkiDanica KragicMårten BjörkmanPublished in: CoRR (2022)
Keyphrases
- generative model
- reinforcement learning
- deep belief networks
- optimal policy
- discriminatively trained
- probabilistic model
- discriminative learning
- discriminative models
- conditional random fields
- mixture model
- em algorithm
- hierarchical hidden markov models
- expectation maximization
- hierarchical models
- training examples
- supervised learning
- semi supervised
- training set
- training samples
- object categories
- learning algorithm
- object detection
- prior knowledge
- reward function
- maximum entropy principle
- bayesian networks