Recycling decision trees in numeric domains

Miroslav Kubat

Research output: Contribution to journal › Article

Abstract

A decision tree's classification performance can drop if the tree is used in a changed context such as different accent in speach recognition. This brittleness can partially be rectified by the use of a cheap second tier implemented as a linear classifier. The transfer of the tree to a novel context is accomplished by re-inducing the second tier, without the need to re-induce the more expensive first tier. Experiments reported in this paper indicate that quick adaptation to the target context can indeed be achieved.

Original languageEnglish
Pages (from-to)195-204
Number of pages10
JournalInformatica (Ljubljana)
Volume24
Issue number2
StatePublished - Jun 1 2000
Externally publishedYes

Fingerprint

Recycling
Brittleness
Decision trees
Numerics
Decision tree
Classifiers
Experiments
Classifier
Target
Experiment
Context

ASJC Scopus subject areas

  • Software

Cite this

Recycling decision trees in numeric domains. / Kubat, Miroslav.

In: Informatica (Ljubljana), Vol. 24, No. 2, 01.06.2000, p. 195-204.

Research output: Contribution to journal › Article

Kubat, M 2000, 'Recycling decision trees in numeric domains', Informatica (Ljubljana), vol. 24, no. 2, pp. 195-204.
Kubat M. Recycling decision trees in numeric domains. Informatica (Ljubljana). 2000 Jun 1;24(2):195-204.
Kubat, Miroslav. / Recycling decision trees in numeric domains. In: Informatica (Ljubljana). 2000 ; Vol. 24, No. 2. pp. 195-204.
@article{3f8e679c0f8e45f3bcea1ad7efbc9e3c,
title = "Recycling decision trees in numeric domains",
abstract = "A decision tree's classification performance can drop if the tree is used in a changed context such as different accent in speach recognition. This brittleness can partially be rectified by the use of a cheap second tier implemented as a linear classifier. The transfer of the tree to a novel context is accomplished by re-inducing the second tier, without the need to re-induce the more expensive first tier. Experiments reported in this paper indicate that quick adaptation to the target context can indeed be achieved.",
author = "Miroslav Kubat",
year = "2000",
month = "6",
day = "1",
language = "English",
volume = "24",
pages = "195--204",
journal = "Informatica (Slovenia)",
issn = "0350-5596",
publisher = "Slovene Society Informatika",
number = "2",

}

TY - JOUR

T1 - Recycling decision trees in numeric domains

AU - Kubat, Miroslav

PY - 2000/6/1

Y1 - 2000/6/1

N2 - A decision tree's classification performance can drop if the tree is used in a changed context such as different accent in speach recognition. This brittleness can partially be rectified by the use of a cheap second tier implemented as a linear classifier. The transfer of the tree to a novel context is accomplished by re-inducing the second tier, without the need to re-induce the more expensive first tier. Experiments reported in this paper indicate that quick adaptation to the target context can indeed be achieved.

AB - A decision tree's classification performance can drop if the tree is used in a changed context such as different accent in speach recognition. This brittleness can partially be rectified by the use of a cheap second tier implemented as a linear classifier. The transfer of the tree to a novel context is accomplished by re-inducing the second tier, without the need to re-induce the more expensive first tier. Experiments reported in this paper indicate that quick adaptation to the target context can indeed be achieved.

UR - http://www.scopus.com/inward/record.url?scp=0034205294&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0034205294&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0034205294

VL - 24

SP - 195

EP - 204

JO - Informatica (Slovenia)

JF - Informatica (Slovenia)

SN - 0350-5596

IS - 2

ER -

Access to Document