Package: multiobjectiveMDP 1.0.0

Anas Mifrani

multiobjectiveMDP: Solution Methods for Multi-Objective Markov Decision Processes

Compendium of the most representative algorithms in print---vector-valued dynamic programming, linear programming, policy iteration, the weighting factor approach---for solving multi-objective Markov decision processes, with or without reward discount, over a finite or infinite horizon. Mifrani, A. (2024) <doi:10.1007/s10479-024-06439-x>; Mifrani, A. & Noll, D. <doi:10.48550/arXiv.2502.13697>; Wakuta, K. (1995) <doi:10.1016/0304-4149(94)00064-Z>.

Authors:Anas Mifrani [aut, cre, cph]

multiobjectiveMDP_1.0.0.tar.gz
multiobjectiveMDP_1.0.0.tar.gz(r-4.7-any)multiobjectiveMDP_1.0.0.tar.gz(r-4.6-any)
multiobjectiveMDP_1.0.0.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
multiobjectiveMDP/json (API)

# Install 'multiobjectiveMDP' in R:

install.packages('multiobjectiveMDP', repos = c('https://cran.r-universe.dev', 'https://cloud.r-project.org'))

On CRAN:

This package does not link to any Github/Gitlab/R-forge repository. No issue tracker or development information is available.

2.00 score 226 downloads 23 exports 51 dependencies

Last updated from:09555c16fb. Checks:4 OK. Indexed: yes.

Target	Result	Time
linux-devel-x86_64	OK	146
source / vignettes	OK	181
linux-release-x86_64	OK	174
wasm-release	OK	145

Exports:are_valid_finite_horizon_rewards are_valid_finite_horizon_transition_probabilities are_valid_infinite_horizon_rewards are_valid_infinite_horizon_transition_probabilities compromise_solution discounted_bellman_operator efficient_subset_sort_prune evaluate_discounted_MMDP_pure_policy evaluate_finite_horizon_MMDP_markov_policy generate_rand_MMDP is_valid_finite_horizon_policy is_valid_infinite_horizon_policy solve_discounted_MDP_policy_iteration solve_discounted_MMDP_linear_programming solve_discounted_MMDP_policy_iteration solve_discounted_MMDP_weighting_factor solve_finite_horizon_MDP_backward_induction solve_finite_horizon_MMDP_backward_induction solve_finite_horizon_MMDP_linear_programming solve_finite_horizon_MMDP_weighting_factor solve_MOLP sum_set value_function_domination_sets

Dependencies:cli codetools cpp11 data.table diagram digest dplyr farver future future.apply generics ggplot2 globals glue gtable isoband KernSmooth labeling lattice lava lifecycle linprog lintools listenv lpSolve magrittr Matrix mco nsga2R numDeriv parallelly pillar pkgconfig pracma prodlim progressr R6 RColorBrewer Rcpp rlang S7 scales shape SQUAREM survival tibble tidyselect utf8 vctrs viridisLite withr

Introduction to multiobjectiveMDP

Rendered fromIntroduction-to-multiobjectiveMDP.Rmdusingknitr::rmarkdown

Last update: 2026-03-06
Started: 2026-03-06

Citation

Readme and manuals

Help Manual

Help page	Topics
Determine whether a numeric list represents a valid reward structure for finite-horizon problems	are_valid_finite_horizon_rewards
Determine whether a numeric list represents a valid transition probability structure for finite-horizon problems	are_valid_finite_horizon_transition_probabilities
Determine whether a numeric list represents a valid reward structure for infinite-horizon problems	are_valid_infinite_horizon_rewards
Determine whether a numeric list represents a valid transition probability structure for infinite-horizon problems	are_valid_infinite_horizon_transition_probabilities
Calculate the compromise solution among a set of objective vectors	compromise_solution
Apply a stationary Bellman-type operator to a vector-valued value function	discounted_bellman_operator
Find the Pareto efficient subset of a set of vectors	efficient_subset_sort_prune
Evaluate a stationary policy in a discounted infinite-horizon multi-objective Markov decision process	evaluate_discounted_MMDP_pure_policy
Evaluate a Markov deterministic policy for a finite-horizon multi-objective Markov decision process	evaluate_finite_horizon_MMDP_markov_policy
Generate a random instance of a multi-objective Markov decision process	generate_rand_MMDP
Determine whether an integer matrix represents a policy for a given class of finite-horizon problems	is_valid_finite_horizon_policy
Determine whether an integer vector represents a stationary policy for a given class of infinite-horizon problems	is_valid_infinite_horizon_policy
Optimize a discounted infinite-horizon Markov decision process through policy iteration	solve_discounted_MDP_policy_iteration
Optimize a discounted infinite-horizon multi-objective Markov decision process through linear programming	solve_discounted_MMDP_linear_programming
Optimize a discounted infinite-horizon multi-objective Markov decision process through policy iteration	solve_discounted_MMDP_policy_iteration
Optimize a discounted infinite-horizon multi-objective Markov decision process through the weighting factor approach	solve_discounted_MMDP_weighting_factor
Solve a standard finite-horizon Markov decision process through dynamic programming	solve_finite_horizon_MDP_backward_induction
Optimize a finite-horizon multi-objective Markov decision process through vector-valued dynamic programming	solve_finite_horizon_MMDP_backward_induction
Optimize a finite-horizon multi-objective Markov decision process through linear programming	solve_finite_horizon_MMDP_linear_programming
Optimize a finite-horizon multi-objective Markov decision process through the weighting factor approach	solve_finite_horizon_MMDP_weighting_factor
Solve a multi-objective linear programming problem by a simplex-type method	solve_MOLP
Calculate the sum set (Minkowski sum) of two or more sets of vectors	sum_set
Compare two or more vector-valued value functions	value_function_domination_sets

Package: multiobjectiveMDP 1.0.0

multiobjectiveMDP: Solution Methods for Multi-Objective Markov Decision Processes

Citation

Readme and manuals

Help Manual

Usage by other packages (reverse dependencies)