A Reinforcement Learning Method with Implicit Critics from a Bystander.
Kao-Shing HwangChi-Wei HsiehWei-Cheng JiangJin-Ling LinPublished in: ISNN (1) (2017)
Keyphrases
- experimental evaluation
- high precision
- detection method
- theoretical analysis
- prior knowledge
- fully automatic
- segmentation method
- classification method
- reinforcement learning
- dynamic programming
- classification accuracy
- support vector machine
- feature set
- cost function
- matching algorithm
- mathematical model
- synthetic data
- mutual information
- detection algorithm
- computationally efficient
- input data
- neural network
- high accuracy
- computational cost
- objective function
- similarity measure