SEAGuL: Sample Efficient Adversarially Guided Learning of Value Functions

Keyphrases

• Value Function Approximation 100%
• Adversarial Samples 100%
• Duality Theory 50%
• Task Balancing 50%
• Adversarial Learning 50%
• Dynamic Nonlinear 50%
• Training Value 50%
• Generalization Error 50%
• Value Function Learning 50%
• Shifting Distribution 50%
• Acrobot 50%

Computer Science

• Generalization Error 33%