A connectionist actor-critic algorithm for faster learning and biological plausibility.
Leonard JohardEmanuele RuffaldiPublished in: ICRA (2014)
Keyphrases
- actor critic
- learning algorithm
- reinforcement learning
- dynamic programming
- gradient method
- np hard
- policy gradient
- cost function
- linear programming
- simulated annealing
- temporal difference
- learning problems
- learning tasks
- monte carlo
- expectation maximization
- neural network
- supervised learning
- prior knowledge
- search space
- computational complexity
- optimal solution
- least squares
- neuro fuzzy
- negative matrix factorization
- approximate dynamic programming
- machine learning