Fast Likelihood Classification

Chulhee Lee, David A. Landgrebe

Research output: Contribution to journalArticle

24 Citations (Scopus)

Abstract

A multistage classification is proposed which reduces the processing time substantially. The proposed classification algorithm consists of several stages, and in each stage likelihood values of classes are calculated and compared. In the proposed multistage classification, if a class has a likelihood value less than a threshold, the class is truncated at that stage as an unlikely class, thus reducing the number of classes for which likelihood values are to be calculated at the next stage. Thus a host of classes can be truncated using a small portion of the total features at early stages, resulting in substantial reduction of computing time. Several truncation criteria are developed and the relationship between thresholds and the error caused by the truncation is investigated. Experiments show that the proposed algorithm reduces the processing time by the factor of 3-7, depending on the number of classes and features, while maintaining essentially the same accuracies.

Original languageEnglish
Pages (from-to)509-517
Number of pages9
JournalIEEE Transactions on Geoscience and Remote Sensing
Volume29
Issue number4
DOIs
Publication statusPublished - 1991 Jan 1

Fingerprint

Processing
thresholds
approximation
experiment
Experiments

All Science Journal Classification (ASJC) codes

  • Earth and Planetary Sciences(all)
  • Electrical and Electronic Engineering
  • Computers in Earth Sciences
  • Geochemistry and Petrology
  • Geophysics

Cite this

Lee, Chulhee ; Landgrebe, David A. / Fast Likelihood Classification. In: IEEE Transactions on Geoscience and Remote Sensing. 1991 ; Vol. 29, No. 4. pp. 509-517.
@article{b67abaeb500145f68c220b6786516168,
title = "Fast Likelihood Classification",
abstract = "A multistage classification is proposed which reduces the processing time substantially. The proposed classification algorithm consists of several stages, and in each stage likelihood values of classes are calculated and compared. In the proposed multistage classification, if a class has a likelihood value less than a threshold, the class is truncated at that stage as an unlikely class, thus reducing the number of classes for which likelihood values are to be calculated at the next stage. Thus a host of classes can be truncated using a small portion of the total features at early stages, resulting in substantial reduction of computing time. Several truncation criteria are developed and the relationship between thresholds and the error caused by the truncation is investigated. Experiments show that the proposed algorithm reduces the processing time by the factor of 3-7, depending on the number of classes and features, while maintaining essentially the same accuracies.",
author = "Chulhee Lee and Landgrebe, {David A.}",
year = "1991",
month = "1",
day = "1",
doi = "10.1109/36.135812",
language = "English",
volume = "29",
pages = "509--517",
journal = "IEEE Transactions on Geoscience and Remote Sensing",
issn = "0196-2892",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "4",

}

Fast Likelihood Classification. / Lee, Chulhee; Landgrebe, David A.

In: IEEE Transactions on Geoscience and Remote Sensing, Vol. 29, No. 4, 01.01.1991, p. 509-517.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Fast Likelihood Classification

AU - Lee, Chulhee

AU - Landgrebe, David A.

PY - 1991/1/1

Y1 - 1991/1/1

N2 - A multistage classification is proposed which reduces the processing time substantially. The proposed classification algorithm consists of several stages, and in each stage likelihood values of classes are calculated and compared. In the proposed multistage classification, if a class has a likelihood value less than a threshold, the class is truncated at that stage as an unlikely class, thus reducing the number of classes for which likelihood values are to be calculated at the next stage. Thus a host of classes can be truncated using a small portion of the total features at early stages, resulting in substantial reduction of computing time. Several truncation criteria are developed and the relationship between thresholds and the error caused by the truncation is investigated. Experiments show that the proposed algorithm reduces the processing time by the factor of 3-7, depending on the number of classes and features, while maintaining essentially the same accuracies.

AB - A multistage classification is proposed which reduces the processing time substantially. The proposed classification algorithm consists of several stages, and in each stage likelihood values of classes are calculated and compared. In the proposed multistage classification, if a class has a likelihood value less than a threshold, the class is truncated at that stage as an unlikely class, thus reducing the number of classes for which likelihood values are to be calculated at the next stage. Thus a host of classes can be truncated using a small portion of the total features at early stages, resulting in substantial reduction of computing time. Several truncation criteria are developed and the relationship between thresholds and the error caused by the truncation is investigated. Experiments show that the proposed algorithm reduces the processing time by the factor of 3-7, depending on the number of classes and features, while maintaining essentially the same accuracies.

UR - http://www.scopus.com/inward/record.url?scp=0026185634&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0026185634&partnerID=8YFLogxK

U2 - 10.1109/36.135812

DO - 10.1109/36.135812

M3 - Article

VL - 29

SP - 509

EP - 517

JO - IEEE Transactions on Geoscience and Remote Sensing

JF - IEEE Transactions on Geoscience and Remote Sensing

SN - 0196-2892

IS - 4

ER -