Posterior Sampling with Delayed Feedback for Reinforcement Learning with Linear Function Approximation

Keyphrases

• Reinforcement Learning 100%
• Delayed Feedback 100%
• Linear Function Approximation 100%
• Posterior Sampling 100%
• Value Function 25%
• Popular 25%
• Blowing 25%
• Real-world Systems 25%
• Sample Complexity 25%
• Taking Action 25%
• Function Space 25%
• Gradient-based 25%
• Computational Cost 25%
• Computational Efficiency 25%
• Langevin Dynamics 25%
• Expected Delay 25%
• Regret 25%
• Optimal Regret 25%
• Approximate Sampling 25%
• Function Approximation 25%
• Sampling Scheme 25%
• Sampling Algorithm 25%
• Order Optimal 25%
• UCB Algorithm 25%
• Reinforcement Learning Problems 25%
• Immediate Feedback 25%
• Noise Perturbation 25%
• Worst-case Regret 25%
• Optimistic Value 25%
• Value-based Algorithms 25%
• Stochastic Delay 25%

Mathematics

• Linear Function 100%
• Worst Case 50%
• Stochastics 50%
• Approximates 50%
• Wide Range 50%
• Computational Cost 50%
• Function Space 50%
• Sampling Scheme 50%
• Real World System 50%
• Approximation Function 50%

Computer Science

• Reinforcement Learning 100%
• Function Approximation 100%
• Linear Function 100%
• Function Value 20%
• Efficient Algorithm 20%
• Learning Problem 20%
• Computational Cost 20%
• Computational Efficiency 20%
• Sampling Scheme 20%
• Sampling Algorithm 20%