Machine Learning Methods in Visualisation for Big Data 2025

Permanent URI for this collection

https://diglib.eg.org/handle/10.2312/3607194

Browse

Browsing Machine Learning Methods in Visualisation for Big Data 2025 by Title

Now showing 1 - 3 of 3
  • Thumbnail Image
    Item
    MLVis 2025: Frontmatter
    (The Eurographics Association, 2025) Archambault, Daniel; Nabney, Ian; Peltonen, Jaakko; Archambault, Daniel; Nabney, Ian; Peltonen, Jaakko
  • Thumbnail Image
    Item
    Neighbour Embeddings: Beyond Visualisation
    (The Eurographics Association, 2025) Lambert, Pierre; Couplet, Edouard; Verleysen, Michel; Lee, John Aldo; Archambault, Daniel; Nabney, Ian; Peltonen, Jaakko
    Machine learning (ML) has brought powerful tools to the visualisation community, particularly through neighbour embeddings (NE). This family of algorithms enables the intuitive visualisation of high dimensional datasets, by representing these in 2- or 3-dimensional spaces. This paper argues that as NE algorithms have progressed and diversified within the visualisation domain, they have matured into powerful yet often simple methods whose potential remains largely underutilised in broader machine learning contexts. This argument is illustrated by showing through two use cases, clustering and data preprocessing before a supervised task, how NE can contribute meaningfully with little additional algorithmic effort.
  • Thumbnail Image
    Item
    Spacewalker: Traversing Representation Spaces for Fast Interactive Exploration and Annotation of Unstructured Data
    (The Eurographics Association, 2025) Heine, Lukas; Hörst, Fabian; Fragemann, Jana; Luijten, Gijs; Egger, Jan; Bahnsen, Fin Hendrik; Sarfraz, M. Saquib; Kleesiek, Jens; Seibold, Constantin; Archambault, Daniel; Nabney, Ian; Peltonen, Jaakko
    In industries such as healthcare, finance, and manufacturing, analysis of unstructured textual data presents significant challenges for analysis and decision making. Uncovering patterns within large-scale corpora and understanding their semantic impact is critical, but depends on domain experts or resource-intensive manual reviews. In response, we introduce Spacewalker, an interactive tool designed to analyze, explore, and annotate data across multiple modalities. It allows users to extract data representations, visualize them in low-dimensional spaces and traverse large datasets either exploratorily or by querying regions of interest. We evaluated Spacewalker through extensive studies, assessing its efficacy in improving data integrity verification and annotation. We show that Spacewalker reduces time and effort compared to traditional methods. The code of this work is publicly available on https://github.com/TIO-IKIM/Spacewalker.