Scoring Levels of Categorical Variables with Heterogeneous Data

Eugene Tuv, George Runger

Research output: Contribution to journalArticle

3 Citations (Scopus)

Abstract

The computationally efficient and flexible method for mapping categorical variables to numeric scores in mixed-type data is described. In categorical variable, categories map to numerical low-dimensional space which allows to discriminate between different nontrivial patterns in data. Clustering uses classification and regression trees (CART) learning engine and has natural mechanism to deal with mixed-type data. Scoring categorical variables enrich data to improve quality of distance-based supervised learning.

Original languageEnglish (US)
Pages (from-to)14-19
Number of pages6
JournalIEEE Intelligent Systems
Volume19
Issue number2
DOIs
StatePublished - Mar 2004

Fingerprint

Supervised learning
Engines

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Electrical and Electronic Engineering
  • Artificial Intelligence

Cite this

Scoring Levels of Categorical Variables with Heterogeneous Data. / Tuv, Eugene; Runger, George.

In: IEEE Intelligent Systems, Vol. 19, No. 2, 03.2004, p. 14-19.

Research output: Contribution to journalArticle

@article{8345df45d7b849bc874bd645d92f740d,
title = "Scoring Levels of Categorical Variables with Heterogeneous Data",
abstract = "The computationally efficient and flexible method for mapping categorical variables to numeric scores in mixed-type data is described. In categorical variable, categories map to numerical low-dimensional space which allows to discriminate between different nontrivial patterns in data. Clustering uses classification and regression trees (CART) learning engine and has natural mechanism to deal with mixed-type data. Scoring categorical variables enrich data to improve quality of distance-based supervised learning.",
author = "Eugene Tuv and George Runger",
year = "2004",
month = "3",
doi = "10.1109/MIS.2004.1274906",
language = "English (US)",
volume = "19",
pages = "14--19",
journal = "IEEE Intelligent Systems and Their Applications",
issn = "1541-1672",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - Scoring Levels of Categorical Variables with Heterogeneous Data

AU - Tuv, Eugene

AU - Runger, George

PY - 2004/3

Y1 - 2004/3

N2 - The computationally efficient and flexible method for mapping categorical variables to numeric scores in mixed-type data is described. In categorical variable, categories map to numerical low-dimensional space which allows to discriminate between different nontrivial patterns in data. Clustering uses classification and regression trees (CART) learning engine and has natural mechanism to deal with mixed-type data. Scoring categorical variables enrich data to improve quality of distance-based supervised learning.

AB - The computationally efficient and flexible method for mapping categorical variables to numeric scores in mixed-type data is described. In categorical variable, categories map to numerical low-dimensional space which allows to discriminate between different nontrivial patterns in data. Clustering uses classification and regression trees (CART) learning engine and has natural mechanism to deal with mixed-type data. Scoring categorical variables enrich data to improve quality of distance-based supervised learning.

UR - http://www.scopus.com/inward/record.url?scp=2142758688&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=2142758688&partnerID=8YFLogxK

U2 - 10.1109/MIS.2004.1274906

DO - 10.1109/MIS.2004.1274906

M3 - Article

AN - SCOPUS:2142758688

VL - 19

SP - 14

EP - 19

JO - IEEE Intelligent Systems and Their Applications

JF - IEEE Intelligent Systems and Their Applications

SN - 1541-1672

IS - 2

ER -