Code covered by the BSD License  

Highlights from
Markov Decision Processes (MDP) Toolbox

• [V,mean_discrepancy]=mdp_eval... mdp_eval_policy_TD_0 Evaluation of the value function, using the TD(0) algorithm
• mdp_LP(P, R, discount) mdp_LP Resolution of discounted MDP with linear programming
• mdp_Q_learning(P, R, discount... mdp_Q_learning Evaluation of the matrix Q, using the Q learning algorithm
• mdp_bellman_operator(P, PR, d... mdp_bellman_operator Applies the Bellman operator on the value function Vprev
• mdp_check(P , R) mdp_check Check if the matrices P and R define a Markov Decision Process
• mdp_check_square_stochastic( ... mdp_check_square_stochastic Check if Z is a square stochastic matrix
• mdp_computePR(P,R) mdp_computePR Computes the reward for the system in one state
• mdp_computePpolicyPRpolicy(P,... mdp_computePpolicyPRpolicy Computes the transition matrix and the reward matrix for a policy
• mdp_eval_policy_iterative(P, ... mdp_eval_policy_iterative Policy evaluation using iteration.
• mdp_eval_policy_matrix(P, R, ... mdp_eval_policy_matrix Evaluation of the value function of a policy
• mdp_eval_policy_optimality(P,... mdp_eval_policy_optimality Eval if near optimum actions exists for each state
• mdp_example_forest (S, r1, r2... mdp_example_forest Generate a Markov Decision Process example based on
• mdp_example_rand (S, A, is_sp... mdp_example_rand Generate a random Markov Decision Process
• mdp_finite_horizon(P, R, disc... mdp_finite_horizon Reolution of finite-horizon MDP with backwards induction
• mdp_policy_iteration(P, R, di... mdp_policy_iteration Resolution of discounted MDP
• mdp_policy_iteration_modified... mdp_policy_iteration_modified Resolution of discounted MDP
• mdp_relative_value_iteration(... mdp_relative_value_iteration Resolution of MDP with average reward
• mdp_silent() mdp_silent Ask for running resolution functions of the MDP Toolbox
• mdp_span(W) mdp_span Returns the span of W
• mdp_value_iteration(P, R, dis... mdp_value_iteration Resolution of discounted MDP with value iteration algorithm
• mdp_value_iterationGS(P, R, d... mdp_value_iterationGS Resolution of discounted MDP with value iteration Gauss-Seidel algorithm
• mdp_value_iteration_bound_ite... mdp_value_iteration_bound_iter Computes a bound for the number of iterations
• mdp_verbose() mdp_verbose Ask for running resolution functions of the MDP Toolbox
• Presentation of MDP toolbox d...
• index_alphabetic.html
• index_category.html
• mdp_LP.html
• mdp_Q_learning.html
• mdp_bellman_operator.html
• mdp_check.html
• mdp_check_square_stochastic.h...
• mdp_computePR.html
• mdp_computePpolicyPRpolicy.ht...
• mdp_eval_policy_TD_0.html
• mdp_eval_policy_iterative.html
• mdp_eval_policy_matrix.html
• mdp_eval_policy_optimality.ht...
• mdp_example_forest.html
• mdp_example_rand.html
• mdp_finite_horizon.html
• mdp_policy_iteration.html
• mdp_policy_iteration_modified...
• mdp_relative_value_iteration....
• mdp_span.html
• mdp_value_iteration.html
• mdp_value_iterationGS.html
• mdp_value_iteration_bound_ite...
• mdp_verbose_silent.html
• View all files