Truncated Stochastic Approximation with Moving Bounds: Convergence

Teo Sharia

doi:10.1007/s11203-014-9093-6

Truncated Stochastic Approximation with Moving Bounds: Convergence

Teo Sharia

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

65 Downloads (Pure)

Abstract

In this paper we consider a wide class of truncated stochastic approximation procedures.
These procedures have three main characteristics: truncations with random moving bounds,
a matrix valued random step-size sequence, and a dynamically changing random regression function.
We establish convergence and consider several examples to illustrate the results

Original language	English
Pages (from-to)	163-179
Number of pages	17
Journal	Statistical Inference for Stochastic Processes
Volume	17
Early online date	7 Mar 2014
DOIs	https://doi.org/10.1007/s11203-014-9093-6
Publication status	Published - Jul 2014

Access to Document

10.1007/s11203-014-9093-6

TrStochApp-TeoSharia-8Accepted author manuscript, 320 KB

http://arxiv.org/abs/1101.0031

Cite this

@article{54c14d75958f4fd081cd5968034f1fd7,

title = "Truncated Stochastic Approximation with Moving Bounds: Convergence",

abstract = "In this paper we consider a wide class of truncated stochastic approximation procedures. These procedures have three main characteristics: truncations with random moving bounds, a matrix valued random step-size sequence, and a dynamically changing random regression function. We establish convergence and consider several examples to illustrate the results",

author = "Teo Sharia",

year = "2014",

month = jul,

doi = "10.1007/s11203-014-9093-6",

language = "English",

volume = "17",

pages = "163--179",

journal = "Statistical Inference for Stochastic Processes",

issn = "1387-0874",

publisher = "Springer Netherlands",

}

TY - JOUR

T1 - Truncated Stochastic Approximation with Moving Bounds

T2 - Convergence

AU - Sharia, Teo

PY - 2014/7

Y1 - 2014/7

N2 - In this paper we consider a wide class of truncated stochastic approximation procedures. These procedures have three main characteristics: truncations with random moving bounds, a matrix valued random step-size sequence, and a dynamically changing random regression function. We establish convergence and consider several examples to illustrate the results

AB - In this paper we consider a wide class of truncated stochastic approximation procedures. These procedures have three main characteristics: truncations with random moving bounds, a matrix valued random step-size sequence, and a dynamically changing random regression function. We establish convergence and consider several examples to illustrate the results

U2 - 10.1007/s11203-014-9093-6

DO - 10.1007/s11203-014-9093-6

M3 - Article

SN - 1387-0874

VL - 17

SP - 163

EP - 179

JO - Statistical Inference for Stochastic Processes

JF - Statistical Inference for Stochastic Processes

ER -