Optimal partitioning of probability distributions under general convex loss functions in selective assembly

Research output: Contribution to journalArticle

9 Citations (Scopus)

Abstract

Selective assembly is an effective approach for improving the quality of a product which is composed of two mating components. This article studies optimal partitioning of the dimensional distributions of the components in selective assembly. It extends previous results for squared error loss function to cover general convex loss functions, including asymmetric convex loss functions. Equations for the optimal partition are derived. Assuming that the density function of the dimensional distribution is log-concave, uniqueness of solutions is established and some properties of the optimal partition are shown. Some numerical results compare the optimal partition with some heuristic partitioning schemes.

Original languageEnglish
Pages (from-to)1545-1560
Number of pages16
JournalCommunications in Statistics - Theory and Methods
Volume40
Issue number9
DOIs
Publication statusPublished - 2011 Jan
Externally publishedYes

Fingerprint

Optimal Partition
Loss Function
Convex function
Partitioning
Probability Distribution
Squared Error Loss Function
Log-concave
Uniqueness of Solutions
Density Function
Heuristics
Cover
Numerical Results

Keywords

  • Asymmetric loss
  • Log-concave density
  • Match gauging
  • Stochastic ordering
  • Variation reduction

ASJC Scopus subject areas

  • Statistics and Probability

Cite this

@article{9bbbc53a1ff041dfaa14815c413956e0,
title = "Optimal partitioning of probability distributions under general convex loss functions in selective assembly",
abstract = "Selective assembly is an effective approach for improving the quality of a product which is composed of two mating components. This article studies optimal partitioning of the dimensional distributions of the components in selective assembly. It extends previous results for squared error loss function to cover general convex loss functions, including asymmetric convex loss functions. Equations for the optimal partition are derived. Assuming that the density function of the dimensional distribution is log-concave, uniqueness of solutions is established and some properties of the optimal partition are shown. Some numerical results compare the optimal partition with some heuristic partitioning schemes.",
keywords = "Asymmetric loss, Log-concave density, Match gauging, Stochastic ordering, Variation reduction",
author = "Shun Matsuura",
year = "2011",
month = "1",
doi = "10.1080/03610920903581002",
language = "English",
volume = "40",
pages = "1545--1560",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "9",

}

TY - JOUR

T1 - Optimal partitioning of probability distributions under general convex loss functions in selective assembly

AU - Matsuura, Shun

PY - 2011/1

Y1 - 2011/1

N2 - Selective assembly is an effective approach for improving the quality of a product which is composed of two mating components. This article studies optimal partitioning of the dimensional distributions of the components in selective assembly. It extends previous results for squared error loss function to cover general convex loss functions, including asymmetric convex loss functions. Equations for the optimal partition are derived. Assuming that the density function of the dimensional distribution is log-concave, uniqueness of solutions is established and some properties of the optimal partition are shown. Some numerical results compare the optimal partition with some heuristic partitioning schemes.

AB - Selective assembly is an effective approach for improving the quality of a product which is composed of two mating components. This article studies optimal partitioning of the dimensional distributions of the components in selective assembly. It extends previous results for squared error loss function to cover general convex loss functions, including asymmetric convex loss functions. Equations for the optimal partition are derived. Assuming that the density function of the dimensional distribution is log-concave, uniqueness of solutions is established and some properties of the optimal partition are shown. Some numerical results compare the optimal partition with some heuristic partitioning schemes.

KW - Asymmetric loss

KW - Log-concave density

KW - Match gauging

KW - Stochastic ordering

KW - Variation reduction

UR - http://www.scopus.com/inward/record.url?scp=79952671529&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=79952671529&partnerID=8YFLogxK

U2 - 10.1080/03610920903581002

DO - 10.1080/03610920903581002

M3 - Article

AN - SCOPUS:79952671529

VL - 40

SP - 1545

EP - 1560

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 9

ER -