Policy Optimization for Markovian Jump Linear Quadratic Control: Gradient Method and Global Convergence

Keyphrases

• Global Convergence 100%
• Linear Quadratic Control 100%
• Gradient Method 100%
• Policy Optimization 100%
• Markovian Jump 100%
• Convergence of the Method 100%
• Markov Jump Linear Systems 66%
• Controller 33%
• Coercivity 33%
• State Feedback Controller 33%
• Mean Square 33%
• Cost Performance 33%
• Quadratic Performance 33%
• Optimal State Feedback 33%
• Non-convexity 33%
• Controller Performance 33%
• Optimization Landscape 33%
• Reinforcement Learning Task 33%
• Quadratic Optimal Control 33%
• Static State Feedback Controller 33%
• Gradient Dominance 33%
• Various Applications 33%
• Useful Properties 33%
• Linear Growth Rate 33%
• Linear Quadratic Control Problem 33%
• Discrete-time Markov Jump Linear Systems 33%
• Policy Gradient Methods 33%

Computer Science

• Global Convergence 100%
• Gradient Method 100%
• Policy Optimization 100%
• Linear Systems 75%
• state-feedback controller 50%
• Reinforcement Learning 25%
• discrete-time 25%
• Optimal Control 25%

Mathematics

• Linear System 100%
• Linear Quadratic Control 100%
• Control Problems 33%
• Optimal Control Theory 33%
• Method Converges 33%
• Mean Square 33%
• Learning Task 33%
• Discrete Time 33%