Beyond the Third Dimension: How Multidimensional Projections and Machine Learning Can Help Each Other

Alexandru Telea

2023

Abstract

Dimensionality reduction (DR) methods, also called projections, are one of the techniques of choice for visually exploring large high-dimensional datasets. In parallel, machine learning (ML) and in particular deep learning applications are one of the most prominent generators of large, high-dimensional, and complex datasets which need visual exploration. As such, it is not surprising that DR methods have been often used to open the black box of ML methods. In this paper, we explore the synergy between developing better DR methods and using them to understand and engineer better ML models. Specific topics covered address selecting suitable DR methods from the wide arena of such available techniques; using ML to create better, faster, and simpler to use direct and inverse projections; extending the projection metaphor to create dense representations of classifiers; and using projections not only to explain, but also to improve, ML models. We end by proposing several high-impact directions for future work that exploit the outlined ML-DR synergy.

Paper Citation

in Harvard Style

Telea A. (2023). Beyond the Third Dimension: How Multidimensional Projections and Machine Learning Can Help Each Other. In Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISIGRAPP, ISBN 978-989-758-634-7, pages 5-16. DOI: 10.5220/0011926400003417

in Bibtex Style

@conference{visigrapp23,
author={Alexandru Telea},
title={Beyond the Third Dimension: How Multidimensional Projections and Machine Learning Can Help Each Other},
booktitle={Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISIGRAPP,},
year={2023},
pages={5-16},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011926400003417},
isbn={978-989-758-634-7},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 18th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications - Volume 5: VISIGRAPP,
TI - Beyond the Third Dimension: How Multidimensional Projections and Machine Learning Can Help Each Other
SN - 978-989-758-634-7
AU - Telea A.
PY - 2023
SP - 5
EP - 16
DO - 10.5220/0011926400003417