Linking Data Separation, Visual Separation, Classifier Performance Using Multidimensional Projections

Bárbara C. Benato*, Alexandre X. Falcão, Alexandru C. Telea

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contributionAcademicpeer-review

Abstract

Understanding how data separation (DS), visual separation (VS), and classifier performance (CP) are related to each other is important for applications in both machine learning and information visualization. A recent study showed that, for a specific machine learning pipeline using a given multidimensional projection technique, high DS leads to high VS and next high CP. However, whether such correlations would stay the same (or not) when using other projection techniques was left open. We fill this gap by evaluating ten projection techniques in a pipeline that uses three contrastive learning methods (SimCLR, SupCon, and their combination) to produce latent spaces and next train and test classifiers for five image datasets of real-world application with human intestinal parasites. Our work identifies two classes of projection techniques – one leading to poor VS and next poor CS regardless of the available DS, and the other showing a good DS-VS-CP correlation. We argue that this last group of projections is a useful instrument in classifier engineering tasks.

Original languageEnglish
Title of host publicationComputer Vision, Imaging and Computer Graphics Theory and Applications - 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics, VISIGRAPP 2023, Revised Selected Papers
EditorsA. Augusto de Sousa, Thomas Bashford-Rogers, Alexis Paljic, Mounia Ziat, Christophe Hurter, Helen Purchase, Petia Radeva, Giovanni Maria Farinella, Kadi Bouatouch
PublisherSpringer
Pages229-255
Number of pages27
ISBN (Electronic)978-3-031-66743-5
ISBN (Print)978-3-031-66742-8
DOIs
Publication statusPublished - 22 Aug 2024
Event18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2023 - Lisbon, Portugal
Duration: 19 Feb 202321 Feb 2023

Publication series

NameCommunications in Computer and Information Science
Volume2103 CCIS
ISSN (Print)1865-0929
ISSN (Electronic)1865-0937

Conference

Conference18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications, VISIGRAPP 2023
Country/TerritoryPortugal
CityLisbon
Period19/02/2321/02/23

Keywords

  • Contrastive learning
  • Data separation
  • Dimensionality reduction algorithms
  • Embedded pseudolabeling
  • Image classification
  • Semi-supervised learning
  • Visual separation

Fingerprint

Dive into the research topics of 'Linking Data Separation, Visual Separation, Classifier Performance Using Multidimensional Projections'. Together they form a unique fingerprint.

Cite this