Policy Iteration is well suited to optimize PageRank
Romain HollandersJean-Charles DelvenneRaphaël M. JungersPublished in: CoRR (2011)
Keyphrases
- policy iteration
- markov decision processes
- model free
- reinforcement learning
- fixed point
- optimal policy
- least squares
- sample path
- markov decision process
- temporal difference
- random walk
- average reward
- policy evaluation
- ranking algorithm
- finite state
- convergence rate
- optimal control
- link analysis
- linear programming
- infinite horizon
- dynamic programming
- discounted reward
- markov decision problems
- web graph
- state space
- link structure
- probabilistic model
- long run
- mathematical model
- support vector