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

Bidirectional classification procedures: double tree and double cluster.

H Hecker1

  • 1Institut für Biometrie(8410), Medizinische Hochschule Hannover, 30623 Hannover, Germany.

Studies in Health Technology and Informatics
|February 24, 2001
PubMed
Summary

This study introduces bidirectional classification, extending prediction rules from CART, RECPAM, and CBR. It uses simultaneous tree or clustering algorithms on predictor (X) and response (Y) variables to find characteristic patterns.

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

Persistent neuropsychiatric impairment in HCV patients despite clearance of the virus?!

Journal of viral hepatitis·2017
Same author

Association of tobacco use and periapical pathosis - a systematic review.

International endodontic journal·2012
Same author

5-ht3 antagonist ondansetron for the treatment of chemotherapy-induced nausea and vomiting in an outpatient population.

International journal of oncology·2011
Same author

Moxifloxacin monotherapy versus ß-lactam mono- or combination therapy in hospitalized patients with community-acquired pneumonia.

The Journal of infection·2011
Same author

Computer-aided classification of colorectal polyps based on vascular patterns: a pilot study.

Endoscopy·2010
Same author

Bovine jugular veins in the pulmonary position in adults -- 5 years' experience with 64 implantations.

The Thoracic and cardiovascular surgeon·2009

Area of Science:

  • Computational statistics
  • Machine learning
  • Data mining

Background:

  • Traditional prediction rules, like those in Classification and Regression Trees (CART), Recursive Partitioning and Amalgamation (RECPAM), or Case-Based Reasoning (CBR), focus on relating predictor variables (X) to a response variable (Y).
  • These methods typically involve unidirectional analysis, limiting the ability to explore reciprocal relationships between predictors and responses.

Purpose of the Study:

  • To extend existing partition-based prediction rule methodologies to bidirectional classification procedures.
  • To develop a novel approach for simultaneously analyzing multivariate predictor variables (X) and a response variable (Y).
  • To detect and characterize inherent patterns and relationships between predictor and response variables within a unified framework.

Main Methods:

Related Experiment Videos

  • Application of tree or clustering algorithms to both predictor (X) and response (Y) variables concurrently.
  • Development of a double partition classification table to represent the bidirectional analysis.
  • Utilizing the generated table to identify characteristic patterns and interrelationships between X and Y values.

Main Results:

  • The bidirectional classification procedure successfully generates a double partition classification table.
  • This table effectively reveals characteristic patterns and associations between multivariate predictor variables and the response variable.
  • The method demonstrates a capability to relate predictor and response values in a more integrated manner than traditional approaches.

Conclusions:

  • Bidirectional classification offers a powerful extension to existing prediction rule methods, enabling simultaneous analysis of predictors and responses.
  • The developed double partition classification table is effective in uncovering complex patterns and relationships.
  • This approach enhances the understanding of how predictor variables collectively influence and relate to response variables.