Best look-alike prediction: Another look at the Bayesian classifier and beyond

Hanmei Sun, Jiming Jiang, Thuan Nguyen, Yihui Luan

Research output: Contribution to journalArticle

Abstract

A criterion of optimality in prediction is proposed that requires the predictor to assume the same type of values as the random variable it is predicting. In the case of categorical responses, the method leads to the Bayesian classifier with a uniform prior. However, the method extends to other cases, such as zero-inflated observations, as well. The method, called best look-alike prediction (BLAP), justifies an “usual practice” from a theoretical standpoint. Application of BLAP to small area estimation is considered. A real-data example is discussed.

Original languageEnglish (US)
Pages (from-to)37-42
Number of pages6
JournalStatistics and Probability Letters
Volume143
DOIs
StatePublished - Dec 1 2018

Fingerprint

Bayesian Classifier
Alike
Prediction
Small Area Estimation
Categorical
Justify
Predictors
Optimality
Random variable
Zero
Classifier

Keywords

  • BLAP
  • Categorical outcome
  • Mixed logistic model
  • Small area estimation
  • Zero-inflated random variable

ASJC Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

Best look-alike prediction : Another look at the Bayesian classifier and beyond. / Sun, Hanmei; Jiang, Jiming; Nguyen, Thuan; Luan, Yihui.

In: Statistics and Probability Letters, Vol. 143, 01.12.2018, p. 37-42.

Research output: Contribution to journalArticle

@article{c4c85f58b33d428db04b8e50a1eba242,
title = "Best look-alike prediction: Another look at the Bayesian classifier and beyond",
abstract = "A criterion of optimality in prediction is proposed that requires the predictor to assume the same type of values as the random variable it is predicting. In the case of categorical responses, the method leads to the Bayesian classifier with a uniform prior. However, the method extends to other cases, such as zero-inflated observations, as well. The method, called best look-alike prediction (BLAP), justifies an “usual practice” from a theoretical standpoint. Application of BLAP to small area estimation is considered. A real-data example is discussed.",
keywords = "BLAP, Categorical outcome, Mixed logistic model, Small area estimation, Zero-inflated random variable",
author = "Hanmei Sun and Jiming Jiang and Thuan Nguyen and Yihui Luan",
year = "2018",
month = "12",
day = "1",
doi = "10.1016/j.spl.2018.07.014",
language = "English (US)",
volume = "143",
pages = "37--42",
journal = "Statistics and Probability Letters",
issn = "0167-7152",
publisher = "Elsevier",

}

TY - JOUR

T1 - Best look-alike prediction

T2 - Another look at the Bayesian classifier and beyond

AU - Sun, Hanmei

AU - Jiang, Jiming

AU - Nguyen, Thuan

AU - Luan, Yihui

PY - 2018/12/1

Y1 - 2018/12/1

N2 - A criterion of optimality in prediction is proposed that requires the predictor to assume the same type of values as the random variable it is predicting. In the case of categorical responses, the method leads to the Bayesian classifier with a uniform prior. However, the method extends to other cases, such as zero-inflated observations, as well. The method, called best look-alike prediction (BLAP), justifies an “usual practice” from a theoretical standpoint. Application of BLAP to small area estimation is considered. A real-data example is discussed.

AB - A criterion of optimality in prediction is proposed that requires the predictor to assume the same type of values as the random variable it is predicting. In the case of categorical responses, the method leads to the Bayesian classifier with a uniform prior. However, the method extends to other cases, such as zero-inflated observations, as well. The method, called best look-alike prediction (BLAP), justifies an “usual practice” from a theoretical standpoint. Application of BLAP to small area estimation is considered. A real-data example is discussed.

KW - BLAP

KW - Categorical outcome

KW - Mixed logistic model

KW - Small area estimation

KW - Zero-inflated random variable

UR - http://www.scopus.com/inward/record.url?scp=85051256347&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85051256347&partnerID=8YFLogxK

U2 - 10.1016/j.spl.2018.07.014

DO - 10.1016/j.spl.2018.07.014

M3 - Article

AN - SCOPUS:85051256347

VL - 143

SP - 37

EP - 42

JO - Statistics and Probability Letters

JF - Statistics and Probability Letters

SN - 0167-7152

ER -