ParaDime: A Framework for Parametric Dimensionality Reduction

dc.contributor.authorHinterreiter, Andreasen_US
dc.contributor.authorHumer, Christinaen_US
dc.contributor.authorKainz, Bernharden_US
dc.contributor.authorStreit, Marcen_US
dc.contributor.editorBujack, Roxanaen_US
dc.contributor.editorArchambault, Danielen_US
dc.contributor.editorSchreck, Tobiasen_US
dc.date.accessioned2023-06-10T06:17:04Z
dc.date.available2023-06-10T06:17:04Z
dc.date.issued2023
dc.description.abstractParaDime is a framework for parametric dimensionality reduction (DR). In parametric DR, neural networks are trained to embed high-dimensional data items in a low-dimensional space while minimizing an objective function. ParaDime builds on the idea that the objective functions of several modern DR techniques result from transformed inter-item relationships. It provides a common interface for specifying these relations and transformations and for defining how they are used within the losses that govern the training process. Through this interface, ParaDime unifies parametric versions of DR techniques such as metric MDS, t-SNE, and UMAP. It allows users to fully customize all aspects of the DR process.We show how this ease of customization makes ParaDime suitable for experimenting with interesting techniques such as hybrid classification/embedding models and supervised DR. This way, ParaDime opens up new possibilities for visualizing high-dimensional data.en_US
dc.description.number3
dc.description.sectionheadersInteraction and Accessibility
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume42
dc.identifier.doi10.1111/cgf.14834
dc.identifier.issn1467-8659
dc.identifier.pages337-348
dc.identifier.pages12 pages
dc.identifier.urihttps://doi.org/10.1111/cgf.14834
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf14834
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.rightsAttribution 4.0 International License
dc.rights.urihttps://creativecommons.org/licenses/by/4.0/
dc.subjectCCS Concepts: Computing methodologies -> Neural networks; Learning latent representations; Human-centered computing -> Visualization systems and tools; Information visualization
dc.subjectComputing methodologies
dc.subjectNeural networks
dc.subjectLearning latent representations
dc.subjectHuman centered computing
dc.subjectVisualization systems and tools
dc.subjectInformation visualization
dc.titleParaDime: A Framework for Parametric Dimensionality Reductionen_US
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
v42i3pp337-348_cgf14834.pdf
Size:
5.61 MB
Format:
Adobe Portable Document Format
Collections