Randomized linear programming solves the Markov decision problem in nearly linear (sometimes sublinear) time

Keyphrases

• Linear Programming 100%
• State Action 100%
• Optimal Policy 100%
• Sublinear Time 100%
• Markov Decision Problem 100%
• Linear Algorithm 33%
• Markov Decision Process 33%
• Stochastic Dynamic Program 33%
• Discount Factor 33%
• Number of States 33%
• Value Policy 33%
• Ergodic 33%
• Primal-dual Updates 33%
• State Transition 33%
• Input Size 33%
• Average Reward 33%
• Discounted Reward 33%
• Exponentiated 33%

Mathematics

• Data Format 33%