Evaluating Autoencoders for Parametric and Invertible Multidimensional Projections
Loading...
Date
2025
Journal Title
Journal ISSN
Volume Title
Publisher
The Eurographics Association
Abstract
Recently, neural networks have gained attention for creating parametric and invertible multidimensional data projections. Parametric projections allow for embedding previously unseen data without recomputing the projection as a whole, while invertible projections enable the generation of new data points. However, these properties have never been explored simultaneously for arbitrary projection methods. We evaluate three autoencoder (AE) architectures for creating parametric and invertible projections. Based on a given projection, we train AEs to learn a mapping into 2D space and an inverse mapping into the original space. We perform a quantitative and qualitative comparison on four datasets of varying dimensionality and pattern complexity using t-SNE. Our results indicate that AEs with a customized loss function can create smoother parametric and inverse projections than feed-forward neural networks while giving users control over the strength of the smoothing effect.
Description
CCS Concepts: Human-centered computing → Visualization; Computing methodologies → Machine learning
@inproceedings{10.2312:eurova.20251099,
booktitle = {EuroVis Workshop on Visual Analytics (EuroVA)},
editor = {Schulz, Hans-Jörg and Villanova, Anna},
title = {{Evaluating Autoencoders for Parametric and Invertible Multidimensional Projections}},
author = {Dennig, Frederik L. and Geyer, Nina and Blumberg, Daniela and Metz, Yannick and Keim, Daniel A.},
year = {2025},
publisher = {The Eurographics Association},
ISSN = {2664-4487},
ISBN = {978-3-03868-283-7},
DOI = {10.2312/eurova.20251099}
}