Description of the GMDP framework and problem representation in the toolbox
For a description, see chapter 2 in the reference user manual.
Quick start
To quickly learn how to use the toolbox, open the file doc/QuickStart.pdf with an appropriate software.
GMDPtoolbox functions
For a detailed description, see the reference user manual.
| GMDP generation |  |
|
gmdp_example_epidemio
| Generates a tiny GMDP epidemics management problem on 3 crop fields |
| |
gmdp_example_grid
| Generates a random GMDP problem on a grid of sites |
| |
| gmdp_example_rand | Generates a random GMDP problem on a random network of sites |
| GMDP resolution |  |
|
gmdp_linear_programming
| Solves a GMDP problem using an approximate linear programming algorithm |
| |
gmdp_policy_iteration_MF
| Solves a GMDP problem using the Mean-Field policy iteration algorithm |
| |
| gmdp_globalize | Translates a GMDP problem into a MDP problem |
| Policy description |  |
|
| gmdp_see_policy | Displays the actions prescribed by a policy |
| |
gmdp_see_policy_tab
| Displays the tables of the actions prescribed by a policy for all explicit neighborhood states |
| |
gmdp_see_policy_graph
| Visualizes the actions prescribed by a policy for a given system state |
| |
gmdp_analyze_policy
| Computes and shows actions repartition of a policy for each site |
| |
gmdp_analyze_policy_neighbor
| Computes and shows actions repartition of a policy for each state of each site |
| Policy evaluation |  |
|
| Mean-Fiel based evaluation |
| |
| gmdp_eval_policy_MF | Evaluates a policy using the Mean-Field approximation |
| |
| |
| Simulation based evaluation |
| |
| gmdp_simulate_policy | Simulates the application of a policy |
| |
gmdp_eval_policy_value
| Estimates the global value evolution over time when applying a policy |
| |
| gmdp_eval_policy_value_site _contribution | Estimates the contributions of all sites to the global value when applying a policy |
| |
gmdp_eval_policy_state_time
| Computes and shows the time spent in each state when applying a policy |
| |
| |
| Evaluation of global value |
| |
gmdp_eval_policy_global_value
| Computes the global expected value function when applying a policy |
| |
gmdp_eval_policy_global_value _state
| Computes the global expected value from an initial global state when applying a policy |
| Utilities |  |
|
| gmdp_check_gmdp | Checks the validity of a GMDP |
| |
| gmdp_check_policy | Checks the validity of a policy for a GMDP |
| |
| gmdp_globalize_local_policy | Computes the global policy from a local policy |
| |
| gmdp_localize_global_policy | If it exists, computes the local policy from a global policy |
| |
| gmdp_see_neighborhood | Visualizes the neighborhood relations between sites |