Establish a discern model of under skewed class distribution

Cien Yun Dai; Chi Jun Lu; Chiu Huin Liao; Chen Chieh Yeh

Establish a discern model of under skewed class distribution

Cien Yun Dai, Chi Jun Lu, Chiu Huin Liao, Chen Chieh Yeh

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Under Skewed Class Distribution Data (SCDD) Real-world applications often involve highly skewed data in decision outcomes. Such a highly skewed class distribution problem, if not properly addressed, would imperil the resulting learning effectiveness. We use one hospital Central Venous Catheter Insertion data for the training data. Among them find CVP-tip attribute are skewed class distribution. The paper uses under-sampling over-sampling and cost-sensitivity to solve SCDD problem. And then use C4.5 algorithms to build up classification model for the central venous catheter insertion. Verify via the expert that use this model can effectively help the medical expert to extract out it about Central Venous Catheter Insertion good knowledge rules.

Original language	English
Title of host publication	IMECS 2007 - International MultiConference of Engineers and Computer Scientists 2007
Pages	801-804
Number of pages	4
Publication status	Published - 2007
Event	International MultiConference of Engineers and Computer Scientists 2007, IMECS 2007 - Kowloon, Hong Kong Duration: 2007 Mar 21 → 2007 Mar 23

Publication series

Name	Lecture Notes in Engineering and Computer Science
ISSN (Print)	2078-0958

Other

Other	International MultiConference of Engineers and Computer Scientists 2007, IMECS 2007
Country/Territory	Hong Kong
City	Kowloon
Period	2007/03/21 → 2007/03/23

Keywords

Cost-sensitivity C4.5 algorithms
Over-sampling
Under-sampling

ASJC Scopus subject areas

Computer Science (miscellaneous)

Cite this

Dai, CY, Lu, CJ, Liao, CH & Yeh, CC 2007, Establish a discern model of under skewed class distribution. in IMECS 2007 - International MultiConference of Engineers and Computer Scientists 2007. Lecture Notes in Engineering and Computer Science, pp. 801-804, International MultiConference of Engineers and Computer Scientists 2007, IMECS 2007, Kowloon, Hong Kong, 2007/03/21.

@inproceedings{345c7785df2f41218f560aee5a7626ed,

title = "Establish a discern model of under skewed class distribution",

abstract = "Under Skewed Class Distribution Data (SCDD) Real-world applications often involve highly skewed data in decision outcomes. Such a highly skewed class distribution problem, if not properly addressed, would imperil the resulting learning effectiveness. We use one hospital Central Venous Catheter Insertion data for the training data. Among them find CVP-tip attribute are skewed class distribution. The paper uses under-sampling over-sampling and cost-sensitivity to solve SCDD problem. And then use C4.5 algorithms to build up classification model for the central venous catheter insertion. Verify via the expert that use this model can effectively help the medical expert to extract out it about Central Venous Catheter Insertion good knowledge rules.",

keywords = "Cost-sensitivity C4.5 algorithms, Over-sampling, Under-sampling",

author = "Dai, {Cien Yun} and Lu, {Chi Jun} and Liao, {Chiu Huin} and Yeh, {Chen Chieh}",

year = "2007",

language = "English",

isbn = "9789889867140",

series = "Lecture Notes in Engineering and Computer Science",

pages = "801--804",

booktitle = "IMECS 2007 - International MultiConference of Engineers and Computer Scientists 2007",

note = "International MultiConference of Engineers and Computer Scientists 2007, IMECS 2007 ; Conference date: 21-03-2007 Through 23-03-2007",

}

TY - GEN

T1 - Establish a discern model of under skewed class distribution

AU - Dai, Cien Yun

AU - Lu, Chi Jun

AU - Liao, Chiu Huin

AU - Yeh, Chen Chieh

PY - 2007

Y1 - 2007

N2 - Under Skewed Class Distribution Data (SCDD) Real-world applications often involve highly skewed data in decision outcomes. Such a highly skewed class distribution problem, if not properly addressed, would imperil the resulting learning effectiveness. We use one hospital Central Venous Catheter Insertion data for the training data. Among them find CVP-tip attribute are skewed class distribution. The paper uses under-sampling over-sampling and cost-sensitivity to solve SCDD problem. And then use C4.5 algorithms to build up classification model for the central venous catheter insertion. Verify via the expert that use this model can effectively help the medical expert to extract out it about Central Venous Catheter Insertion good knowledge rules.

AB - Under Skewed Class Distribution Data (SCDD) Real-world applications often involve highly skewed data in decision outcomes. Such a highly skewed class distribution problem, if not properly addressed, would imperil the resulting learning effectiveness. We use one hospital Central Venous Catheter Insertion data for the training data. Among them find CVP-tip attribute are skewed class distribution. The paper uses under-sampling over-sampling and cost-sensitivity to solve SCDD problem. And then use C4.5 algorithms to build up classification model for the central venous catheter insertion. Verify via the expert that use this model can effectively help the medical expert to extract out it about Central Venous Catheter Insertion good knowledge rules.

KW - Cost-sensitivity C4.5 algorithms

KW - Over-sampling

KW - Under-sampling

UR - http://www.scopus.com/inward/record.url?scp=84888334248&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84888334248&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:84888334248

SN - 9789889867140

T3 - Lecture Notes in Engineering and Computer Science

SP - 801

EP - 804

BT - IMECS 2007 - International MultiConference of Engineers and Computer Scientists 2007

T2 - International MultiConference of Engineers and Computer Scientists 2007, IMECS 2007

Y2 - 21 March 2007 through 23 March 2007

ER -

Establish a discern model of under skewed class distribution

Abstract

Publication series

Other

Keywords

ASJC Scopus subject areas

Other files and links

Fingerprint

Cite this