Applying statistical generalization to determine search direction for reinforcement learning of movement primitives.
Bojan NemecDenis ForteRok VugaMinija TamosiunaiteFlorentin WörgötterAles UdePublished in: Humanoids (2012)
Keyphrases
- reinforcement learning
- search direction
- convergence analysis
- optimal solution
- linear programming problems
- step size
- nonlinear optimization
- state space
- function approximation
- high level
- temporal difference
- primal dual
- optimal policy
- genetic programming
- denoising
- dynamic programming
- cost function
- search algorithm
- genetic algorithm