Multiply Accelerated Value Iteration for Non-Symmetric Affine Fixed Point Problems and application to Markov Decision Processes - Archive ouverte HAL Access content directly
Journal Articles SIAM Journal on Matrix Analysis and Applications Year : 2022

Multiply Accelerated Value Iteration for Non-Symmetric Affine Fixed Point Problems and application to Markov Decision Processes

(1) , (1) , (2) , (1)
1
2
Marianne Akian
  • Function : Author
Stéphane Gaubert
Zheng Qu
  • Function : Author
  • PersonId : 1052547
Omar Saadi
  • Function : Author

Abstract

We analyze a modified version of Nesterov accelerated gradient algorithm, which applies to affine fixed point problems with non self-adjoint matrices, such as the ones appearing in the theory of Markov decision processes with discounted or mean payoff criteria. We characterize the spectra of matrices for which this algorithm does converge with an accelerated asymptotic rate. We also introduce a $d$th-order algorithm, and show that it yields a multiply accelerated rate under more demanding conditions on the spectrum. We subsequently apply these methods to develop accelerated schemes for non-linear fixed point problems arising from Markov decision processes. This is illustrated by numerical experiments.

Dates and versions

hal-03059718 , version 1 (13-12-2020)

Identifiers

Cite

Marianne Akian, Stéphane Gaubert, Zheng Qu, Omar Saadi. Multiply Accelerated Value Iteration for Non-Symmetric Affine Fixed Point Problems and application to Markov Decision Processes. SIAM Journal on Matrix Analysis and Applications, 2022, 43 (1), ⟨10.1137/20M1367192⟩. ⟨hal-03059718⟩
77 View
0 Download

Altmetric

Share

Gmail Facebook Twitter LinkedIn More