Continuous state dynamic programming via nonexpansive approximation

John Stachurski

doi:10.1007/s10614-007-9111-5

Continuous state dynamic programming via nonexpansive approximation

John Stachurski^*

^*Corresponding author for this work

Research output: Contribution to journal › Article › peer-review

19 Citations (Scopus)

Abstract

This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm.

Original language	English
Pages (from-to)	141-160
Number of pages	20
Journal	Computational Economics
Volume	31
Issue number	2
DOIs	https://doi.org/10.1007/s10614-007-9111-5
Publication status	Published - Mar 2008
Externally published	Yes

Access to Document

10.1007/s10614-007-9111-5

Cite this

@article{7e65cca4a308492c8bde724e59db4a23,

title = "Continuous state dynamic programming via nonexpansive approximation",

abstract = "This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm.",

keywords = "Nonexpansive approximation, Numerical dynamic programming",

author = "John Stachurski",

year = "2008",

month = mar,

doi = "10.1007/s10614-007-9111-5",

language = "English",

volume = "31",

pages = "141--160",

journal = "Computational Economics",

issn = "0927-7099",

publisher = "Springer Netherlands",

number = "2",

}

TY - JOUR

T1 - Continuous state dynamic programming via nonexpansive approximation

AU - Stachurski, John

PY - 2008/3

Y1 - 2008/3

N2 - This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm.

AB - This paper studies fitted value iteration for continuous state numerical dynamic programming using nonexpansive function approximators. A number of approximation schemes are discussed. The main contribution is to provide error bounds for approximate optimal policies generated by the value iteration algorithm.

KW - Nonexpansive approximation

KW - Numerical dynamic programming

UR - http://www.scopus.com/inward/record.url?scp=40549131285&partnerID=8YFLogxK

U2 - 10.1007/s10614-007-9111-5

DO - 10.1007/s10614-007-9111-5

M3 - Article

SN - 0927-7099

VL - 31

SP - 141

EP - 160

JO - Computational Economics

JF - Computational Economics

IS - 2

ER -

Continuous state dynamic programming via nonexpansive approximation

Abstract

Access to Document

Other files and links

Fingerprint

Cite this