Trajectory Following Dynamic Programming algorithms without finite support assumptions - École des Ponts ParisTech Access content directly
Journal Articles Journal of Convex Analysis Year : 2023

Trajectory Following Dynamic Programming algorithms without finite support assumptions

Maël Forcier
  • Function : Author
  • PersonId : 1129375
Vincent Leclère
  • Function : Author
  • PersonId : 1008753

Abstract

We introduce a class of algorithms, called Trajectory Following Dynamic Programming (TFDP) algorithms, that iteratively refines approximations of cost-to-go functions of multistage stochastic problems with independent random variables. This framework encompasses most variants of the Stochastic Dual Dynamic Programming algorithm. Leveraging a Lipschitz assumption on the expected cost-to-go functions, we provide a new convergence and complexity proof that allows random variables with non-finitely supported distributions. In particular, this leads to new complexity results for numerous known algorithms. Further, we detail how TFDP algorithms can be implemented without the finite support assumption, either through approximations or exact computations.
Fichier principal
Vignette du fichier
TFDP_final.pdf (722.35 Ko) Télécharger le fichier
Origin : Files produced by the author(s)

Dates and versions

hal-03683697 , version 1 (31-05-2022)
hal-03683697 , version 2 (04-09-2023)

Identifiers

  • HAL Id : hal-03683697 , version 2

Cite

Maël Forcier, Vincent Leclère. Trajectory Following Dynamic Programming algorithms without finite support assumptions. Journal of Convex Analysis, 2023, 30 (3), pp.951-999. ⟨hal-03683697v2⟩
58 View
153 Download

Share

Gmail Facebook X LinkedIn More