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

A coupling procedure for the discrimination of mixed data.

K D Wernecke1

  • 1Humboldt-University of Berlin, Charite-Eye-Clinic, Germany.

Biometrics
|June 1, 1992
PubMed
Summary

This study introduces a new decision rule for combining diverse data structures and scales. The method, validated with medical data, outperforms linear discriminant analysis in reducing error rates.

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

Why does a point of care guided transfusion algorithm not improve blood loss and transfusion practice in patients undergoing high-risk cardiac surgery? A prospective randomized controlled pilot study.

BMC anesthesiology·2019
Same author

The new agreement of the international RIGA consensus conference on nasal airway function tests.

Rhinology·2018
Same author

Feasibility of noise reduction by a modification in ICU environment.

Physiological measurement·2016
Same author

Long-term effect of computer-assisted decision support for antibiotic treatment in critically ill patients: a prospective 'before/after' cohort study.

BMJ open·2014
Same author

Preoperative depression and hospital length of stay in surgical patients.

Minerva anestesiologica·2013
Same author

Monitoring depth of anaesthesia in a randomized trial decreases the rate of postoperative delirium but not postoperative cognitive dysfunction.

British journal of anaesthesia·2013

Area of Science:

  • Statistics
  • Machine Learning
  • Medical Diagnostics

Background:

  • Combining data with varying structures and measurement scales presents a challenge in statistical analysis.
  • Existing methods often have limitations regarding the number and type of features, particularly categorical ones.

Purpose of the Study:

  • To present a novel procedure for coupling different discriminators to a common decision rule.
  • To enable joint analysis of data with heterogeneous structures and scales without stringent feature restrictions.
  • To demonstrate the procedure's utility in medical diagnostics through comparative analysis.

Main Methods:

  • A new decision rule is proposed, utilizing allocation vectors to couple diverse discriminators.
  • The method accommodates data with different structures and/or scales of measurement.
  • A rigorous cross-validation process is integrated to ensure the reliability of the results.

Main Results:

  • The procedure effectively handles data of mixed types and scales, including numerous categorical features.
  • Examples in medical diagnostics show the proposed method's superior performance.
  • Error rates obtained from the new procedure are lower compared to traditional linear discriminant analysis.

Conclusions:

  • The developed procedure offers a flexible and effective approach for integrating diverse datasets.
  • It provides a significant improvement over existing methods like linear discriminant analysis, especially in complex diagnostic scenarios.
  • The cross-validation ensures robust and reliable outcomes for practical applications.

Related Experiment Videos