Constrained Variational Policy Optimization for Safe Reinforcement Learning

Keyphrases

• Variational 100%
• Policy Optimization 100%
• Safe Reinforcement Learning 100%
• Convex Optimization 66%
• Variational Distribution 66%
• Region-based 33%
• Nonparametric 33%
• Learning Objectives 33%
• Sampling Efficiency 33%
• Satisfaction Performance 33%
• Training Performance 33%
• Expectation Maximization 33%
• Safety-critical Applications 33%
• Robotic Tasks 33%
• Supervised Learning 33%
• Probabilistic Inference 33%
• Trust Region 33%
• Constraint Satisfaction 33%
• Optimality Guarantee 33%
• Primal-dual 33%
• Learning Phase 33%
• Policy Parameters 33%
• Policy Learning 33%
• Reinforcement Learning Problems 33%
• Stable Training 33%

Computer Science

• Reinforcement Learning 100%
• Optimization Policy 100%
• Convex Optimization 66%
• Learning Problem 33%
• Critical Application 33%
• Learning Phase 33%
• Primal-Dual 33%
• Supervised Learning 33%
• Constraint Satisfaction Problems 33%
• Probabilistic Inference 33%