Application of multivariable optimal discriminant analysis in general internal medicine

Paul R. Yarnold, Robert C. Soltysik, Wayne C. McCormick, Robert Burns, Elizabeth H.B. Lin, Terry Bush, Gary J. Martin

Research output: Contribution to journalArticle

5 Citations (Scopus)

Abstract

OBJECTIVE: To illustrate the use of multivariable optimal discriminant analysis (MultiODA). DESIGN: Data from four previously published studies were reanalyzed using MultiODA. The original analysis was Fisher's linear discriminant analysis (FLDA) for two studies and logistic regression analysis (LRA) for two studies. MEASUREMENTS AND MAIN RESULTS: In Study 1, FLDA achieved an overall percentage accuracy in classification (PAC) for the training sample of 69.9%, compared with 73.5% for MultiODA. In Study 2, the LRA model required three attributes to achieve a 76.1% overall PAC for the training sample and a 79.4% overall PAC for the hold-out sample. Using only two attributes, the MultiODA model achieved similar values. In Study 3, the FLDA model achieved an overall PAC of 82.5%, compared with 87.5% for the MultiODA model. In Study 4, MultiODA identified a two-attribute model that achieved a 93.3% overall training PAC, when an LRA model could not be developed. CONCLUSIONS: MultiODA identified: a superior training model (Study 1); a more parsimonious model that achieved superior overall training and identical hold-out PAC (Study 2); a model that achieved a higher hold-out PAC (Study 3); and a two-attribute model that achieved a relatively high PAC when a multivariable LRA model could not be obtained (Study 4). These findings suggest that MultiODA has the potential to improve the accuracy of predictions made in general internal medicine research.

Original languageEnglish (US)
Pages (from-to)601-606
Number of pages6
JournalJournal of General Internal Medicine
Volume10
Issue number11
DOIs
StatePublished - Nov 1 1995

Fingerprint

Discriminant Analysis
Internal Medicine
Logistic Models
Regression Analysis

All Science Journal Classification (ASJC) codes

  • Internal Medicine

Cite this

Yarnold, P. R., Soltysik, R. C., McCormick, W. C., Burns, R., Lin, E. H. B., Bush, T., & Martin, G. J. (1995). Application of multivariable optimal discriminant analysis in general internal medicine. Journal of General Internal Medicine, 10(11), 601-606. https://doi.org/10.1007/BF02602743

Application of multivariable optimal discriminant analysis in general internal medicine. / Yarnold, Paul R.; Soltysik, Robert C.; McCormick, Wayne C.; Burns, Robert; Lin, Elizabeth H.B.; Bush, Terry; Martin, Gary J.

In: Journal of General Internal Medicine, Vol. 10, No. 11, 01.11.1995, p. 601-606.

Research output: Contribution to journalArticle

Yarnold, PR, Soltysik, RC, McCormick, WC, Burns, R, Lin, EHB, Bush, T & Martin, GJ 1995, 'Application of multivariable optimal discriminant analysis in general internal medicine', Journal of General Internal Medicine, vol. 10, no. 11, pp. 601-606. https://doi.org/10.1007/BF02602743
Yarnold, Paul R. ; Soltysik, Robert C. ; McCormick, Wayne C. ; Burns, Robert ; Lin, Elizabeth H.B. ; Bush, Terry ; Martin, Gary J. / Application of multivariable optimal discriminant analysis in general internal medicine. In: Journal of General Internal Medicine. 1995 ; Vol. 10, No. 11. pp. 601-606.
@article{14110dcc796c4f6580a8206c12fbce15,
title = "Application of multivariable optimal discriminant analysis in general internal medicine",
abstract = "OBJECTIVE: To illustrate the use of multivariable optimal discriminant analysis (MultiODA). DESIGN: Data from four previously published studies were reanalyzed using MultiODA. The original analysis was Fisher's linear discriminant analysis (FLDA) for two studies and logistic regression analysis (LRA) for two studies. MEASUREMENTS AND MAIN RESULTS: In Study 1, FLDA achieved an overall percentage accuracy in classification (PAC) for the training sample of 69.9{\%}, compared with 73.5{\%} for MultiODA. In Study 2, the LRA model required three attributes to achieve a 76.1{\%} overall PAC for the training sample and a 79.4{\%} overall PAC for the hold-out sample. Using only two attributes, the MultiODA model achieved similar values. In Study 3, the FLDA model achieved an overall PAC of 82.5{\%}, compared with 87.5{\%} for the MultiODA model. In Study 4, MultiODA identified a two-attribute model that achieved a 93.3{\%} overall training PAC, when an LRA model could not be developed. CONCLUSIONS: MultiODA identified: a superior training model (Study 1); a more parsimonious model that achieved superior overall training and identical hold-out PAC (Study 2); a model that achieved a higher hold-out PAC (Study 3); and a two-attribute model that achieved a relatively high PAC when a multivariable LRA model could not be obtained (Study 4). These findings suggest that MultiODA has the potential to improve the accuracy of predictions made in general internal medicine research.",
author = "Yarnold, {Paul R.} and Soltysik, {Robert C.} and McCormick, {Wayne C.} and Robert Burns and Lin, {Elizabeth H.B.} and Terry Bush and Martin, {Gary J.}",
year = "1995",
month = "11",
day = "1",
doi = "10.1007/BF02602743",
language = "English (US)",
volume = "10",
pages = "601--606",
journal = "Journal of General Internal Medicine",
issn = "0884-8734",
publisher = "Springer New York",
number = "11",

}

TY - JOUR

T1 - Application of multivariable optimal discriminant analysis in general internal medicine

AU - Yarnold, Paul R.

AU - Soltysik, Robert C.

AU - McCormick, Wayne C.

AU - Burns, Robert

AU - Lin, Elizabeth H.B.

AU - Bush, Terry

AU - Martin, Gary J.

PY - 1995/11/1

Y1 - 1995/11/1

N2 - OBJECTIVE: To illustrate the use of multivariable optimal discriminant analysis (MultiODA). DESIGN: Data from four previously published studies were reanalyzed using MultiODA. The original analysis was Fisher's linear discriminant analysis (FLDA) for two studies and logistic regression analysis (LRA) for two studies. MEASUREMENTS AND MAIN RESULTS: In Study 1, FLDA achieved an overall percentage accuracy in classification (PAC) for the training sample of 69.9%, compared with 73.5% for MultiODA. In Study 2, the LRA model required three attributes to achieve a 76.1% overall PAC for the training sample and a 79.4% overall PAC for the hold-out sample. Using only two attributes, the MultiODA model achieved similar values. In Study 3, the FLDA model achieved an overall PAC of 82.5%, compared with 87.5% for the MultiODA model. In Study 4, MultiODA identified a two-attribute model that achieved a 93.3% overall training PAC, when an LRA model could not be developed. CONCLUSIONS: MultiODA identified: a superior training model (Study 1); a more parsimonious model that achieved superior overall training and identical hold-out PAC (Study 2); a model that achieved a higher hold-out PAC (Study 3); and a two-attribute model that achieved a relatively high PAC when a multivariable LRA model could not be obtained (Study 4). These findings suggest that MultiODA has the potential to improve the accuracy of predictions made in general internal medicine research.

AB - OBJECTIVE: To illustrate the use of multivariable optimal discriminant analysis (MultiODA). DESIGN: Data from four previously published studies were reanalyzed using MultiODA. The original analysis was Fisher's linear discriminant analysis (FLDA) for two studies and logistic regression analysis (LRA) for two studies. MEASUREMENTS AND MAIN RESULTS: In Study 1, FLDA achieved an overall percentage accuracy in classification (PAC) for the training sample of 69.9%, compared with 73.5% for MultiODA. In Study 2, the LRA model required three attributes to achieve a 76.1% overall PAC for the training sample and a 79.4% overall PAC for the hold-out sample. Using only two attributes, the MultiODA model achieved similar values. In Study 3, the FLDA model achieved an overall PAC of 82.5%, compared with 87.5% for the MultiODA model. In Study 4, MultiODA identified a two-attribute model that achieved a 93.3% overall training PAC, when an LRA model could not be developed. CONCLUSIONS: MultiODA identified: a superior training model (Study 1); a more parsimonious model that achieved superior overall training and identical hold-out PAC (Study 2); a model that achieved a higher hold-out PAC (Study 3); and a two-attribute model that achieved a relatively high PAC when a multivariable LRA model could not be obtained (Study 4). These findings suggest that MultiODA has the potential to improve the accuracy of predictions made in general internal medicine research.

UR - http://www.scopus.com/inward/record.url?scp=0028785103&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0028785103&partnerID=8YFLogxK

U2 - 10.1007/BF02602743

DO - 10.1007/BF02602743

M3 - Article

VL - 10

SP - 601

EP - 606

JO - Journal of General Internal Medicine

JF - Journal of General Internal Medicine

SN - 0884-8734

IS - 11

ER -