SDSRA: A Skill-Driven Skill-Recombination Algorithm for Efficient Policy Learning.
Eric Hanchen JiangAndrew LizarragaPublished in: Tiny Papers @ ICLR (2024)
Keyphrases
- learning algorithm
- dynamic programming
- incremental learning
- detection algorithm
- computationally efficient
- segmentation algorithm
- single pass
- computational complexity
- search space
- optimal solution
- experimental evaluation
- preprocessing
- supervised learning
- high accuracy
- worst case
- learning rules
- reinforcement learning
- optimization algorithm
- theoretical analysis
- particle swarm optimization
- linear programming
- cost function
- learning speed
- motor skills
- learned models
- policy search
- learning tasks
- computational cost
- active learning
- k means
- learning process
- image segmentation