Dual REPS: A Generalization of Relative Entropy Policy Search Exploiting Bad Experiences.
Adria ColomeCarme TorrasPublished in: IEEE Trans. Robotics (2017)
Keyphrases
- relative entropy
- policy search
- information theoretic
- reinforcement learning
- mutual information
- information theory
- log likelihood
- reinforcement learning algorithms
- covariance matrix
- dynamic programming
- continuous state
- mahalanobis distance
- bregman divergences
- maximum entropy
- state space
- markov chain
- reward function
- least squares
- knn
- kullback leibler divergence
- data points