SDSRA: A Skill-Driven Skill-Recombination Algorithm for Efficient Policy Learning.
Eric H. JiangAndrew LizarragaPublished in: CoRR (2023)
Keyphrases
- learning algorithm
- single pass
- learning phase
- learning process
- incremental learning
- high efficiency
- preprocessing
- optimization algorithm
- computational cost
- theoretical analysis
- learning speed
- detection algorithm
- k means
- dynamic programming
- experimental evaluation
- high accuracy
- computationally efficient
- highly efficient
- recognition algorithm
- model free
- expectation maximization
- linear programming
- simulated annealing
- cost function
- active learning
- control policy
- reinforcement learning
- motor skills
- statistical queries
- particle swarm optimization
- convergence rate
- matching algorithm
- segmentation algorithm
- online learning
- significant improvement
- computational complexity
- support vector
- similarity measure
- feature selection