Derivation of Monotone Decision Models from Non-Monotone Data

H.A.M. Daniëls, M.V. Velikova

Research output: Working paperDiscussion paperOther research output

310 Downloads (Pure)

Abstract

The objective of data mining is the extraction of knowledge from databases. In practice, one often encounters difficulties with models that are constructed purely by search, without incorporation of knowledge about the domain of application.In economic decision making such as credit loan approval or risk analysis, one often requires models that are monotone with respect to the decision variables involved.If the model is obtained by a blind search through the data, it does mostly not have this property even if the underlying database is monotone.In this paper, we present methods to enforce monotonicity of decision models.We propose measures to express the degree of monotonicity of the data and an algorithm to make data sets monotone.In addition, it is shown that monotone decision trees derived from cleaned data perform better compared to trees derived from raw data.
Original languageEnglish
Place of PublicationTilburg
PublisherOperations research
Number of pages16
Volume2003-30
Publication statusPublished - 2003

Publication series

NameCentER Discussion Paper
Volume2003-30

Fingerprint

Risk analysis
Decision trees
Data mining
Decision making
Economics

Keywords

  • decision models
  • knowledge
  • decision theory
  • operational research
  • data mining

Cite this

Daniëls, H. A. M., & Velikova, M. V. (2003). Derivation of Monotone Decision Models from Non-Monotone Data. (CentER Discussion Paper; Vol. 2003-30). Tilburg: Operations research.
Daniëls, H.A.M. ; Velikova, M.V. / Derivation of Monotone Decision Models from Non-Monotone Data. Tilburg : Operations research, 2003. (CentER Discussion Paper).
@techreport{d52d436a1736409ab1fc01de8c30b0b5,
title = "Derivation of Monotone Decision Models from Non-Monotone Data",
abstract = "The objective of data mining is the extraction of knowledge from databases. In practice, one often encounters difficulties with models that are constructed purely by search, without incorporation of knowledge about the domain of application.In economic decision making such as credit loan approval or risk analysis, one often requires models that are monotone with respect to the decision variables involved.If the model is obtained by a blind search through the data, it does mostly not have this property even if the underlying database is monotone.In this paper, we present methods to enforce monotonicity of decision models.We propose measures to express the degree of monotonicity of the data and an algorithm to make data sets monotone.In addition, it is shown that monotone decision trees derived from cleaned data perform better compared to trees derived from raw data.",
keywords = "decision models, knowledge, decision theory, operational research, data mining",
author = "H.A.M. Dani{\"e}ls and M.V. Velikova",
note = "Subsequently published in IEEE Transactions on Systems, Man and Cybernetics - Part C (2006) Pagination: 16",
year = "2003",
language = "English",
volume = "2003-30",
series = "CentER Discussion Paper",
publisher = "Operations research",
type = "WorkingPaper",
institution = "Operations research",

}

Daniëls, HAM & Velikova, MV 2003 'Derivation of Monotone Decision Models from Non-Monotone Data' CentER Discussion Paper, vol. 2003-30, Operations research, Tilburg.

Derivation of Monotone Decision Models from Non-Monotone Data. / Daniëls, H.A.M.; Velikova, M.V.

Tilburg : Operations research, 2003. (CentER Discussion Paper; Vol. 2003-30).

Research output: Working paperDiscussion paperOther research output

TY - UNPB

T1 - Derivation of Monotone Decision Models from Non-Monotone Data

AU - Daniëls, H.A.M.

AU - Velikova, M.V.

N1 - Subsequently published in IEEE Transactions on Systems, Man and Cybernetics - Part C (2006) Pagination: 16

PY - 2003

Y1 - 2003

N2 - The objective of data mining is the extraction of knowledge from databases. In practice, one often encounters difficulties with models that are constructed purely by search, without incorporation of knowledge about the domain of application.In economic decision making such as credit loan approval or risk analysis, one often requires models that are monotone with respect to the decision variables involved.If the model is obtained by a blind search through the data, it does mostly not have this property even if the underlying database is monotone.In this paper, we present methods to enforce monotonicity of decision models.We propose measures to express the degree of monotonicity of the data and an algorithm to make data sets monotone.In addition, it is shown that monotone decision trees derived from cleaned data perform better compared to trees derived from raw data.

AB - The objective of data mining is the extraction of knowledge from databases. In practice, one often encounters difficulties with models that are constructed purely by search, without incorporation of knowledge about the domain of application.In economic decision making such as credit loan approval or risk analysis, one often requires models that are monotone with respect to the decision variables involved.If the model is obtained by a blind search through the data, it does mostly not have this property even if the underlying database is monotone.In this paper, we present methods to enforce monotonicity of decision models.We propose measures to express the degree of monotonicity of the data and an algorithm to make data sets monotone.In addition, it is shown that monotone decision trees derived from cleaned data perform better compared to trees derived from raw data.

KW - decision models

KW - knowledge

KW - decision theory

KW - operational research

KW - data mining

M3 - Discussion paper

VL - 2003-30

T3 - CentER Discussion Paper

BT - Derivation of Monotone Decision Models from Non-Monotone Data

PB - Operations research

CY - Tilburg

ER -

Daniëls HAM, Velikova MV. Derivation of Monotone Decision Models from Non-Monotone Data. Tilburg: Operations research. 2003. (CentER Discussion Paper).