Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Generalized covariance-adjusted discriminants: perspective and application

X M Tu1, J Kowalski, J Randall

  • 1Department of Statistics, University of Pittsburgh, Pennsylvania 16260, USA.

Biometrics
|September 18, 1997
PubMed
Summary

This study introduces a generalized covariance-adjusted model for diagnostic marker analysis, relaxing assumptions and enabling variable selection. The new method improves classification accuracy for diagnostic markers in complex patient groups.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Therapeutic drug monitoring of mood stabilizers in Medicaid patients with bipolar disorder.

The American journal of psychiatry·1999
Same author

Relationship between social rhythms and mood in patients with rapid cycling bipolar disorder.

Psychiatry research·1999
Same author

Diagnostic certainty of a voluntary bipolar disorder case registry.

Journal of affective disorders·1999
Same author

Cerebral glucose metabolic response to combined total sleep deprivation and antidepressant treatment in geriatric depression.

The American journal of psychiatry·1999
Same author

Insulin-like growth factor 1 and growth hormone binding protein in depression: a preliminary communication.

Journal of psychiatric research·1999
Same author

Consensus criteria for traumatic grief. A preliminary empirical test.

The British journal of psychiatry : the journal of mental science·1999

Area of Science:

  • Biostatistics
  • Medical Informatics
  • Psychiatric Research

Background:

  • Discriminant analysis for diagnostic markers requires covariate adjustment.
  • Traditional methods are limited by assumptions of linearity and normal distributions.
  • Lack of variable selection methods hinders covariance-adjusted models.

Purpose of the Study:

  • To generalize covariance-adjusted models for discriminant analysis.
  • To relax distributional assumptions and allow nonlinear covariates.
  • To develop variable selection methods for generalized models.

Main Methods:

  • Generalized normal and logistic models were developed.
  • Exact and asymptotic tests for variable selection were derived.
  • Methodology validated using simulated and psychiatric study data.

Related Experiment Videos

Main Results:

  • The generalized models accommodate nonlinear covariates and non-normal feature vectors.
  • Effective variable selection was achieved within the new framework.
  • The approach demonstrated utility in a psychiatric study on anxiety disorders.

Conclusions:

  • The proposed generalized models offer a more flexible and powerful approach to covariate adjustment in discriminant analysis.
  • The developed variable selection methods enhance the practical application of these models.
  • This work provides a valuable tool for analyzing diagnostic markers in heterogeneous patient populations.