Expected Model Change

Expected Model Change is an acquisition function in active learning that quantifies the anticipated parameter update of a model upon training on a candidate unlabeled instance, distinguishing itself from other selection criteria by estimating the magnitude of change in model parameters as a proxy for informativeness.

23

Publications

1.2K

Citations

85

Authors

52

Institutions

Publications per year

2017–2026

9

Authors

85

  1. JD
    Joachim Denzler

    Friedrich Schiller University Jena, Germany

    3 publications · 86 citations

    3

    h-index

  2. ER
    Erik Rodner

    Friedrich Schiller University Jena, Germany

    3 publications · 86 citations

    3

    h-index

  3. CK
    Christoph Käding

    Friedrich Schiller University Jena, Germany

    3 publications · 86 citations

    3

    h-index

  4. AF
    Alexander Freytag

    Friedrich Schiller University Jena, Germany

    3 publications · 86 citations

    3

    h-index

  5. WC
    Wenbin Cai

    Beijing University of Technology, China

    2 publications · 274 citations

    2

    h-index

1–5 of 85

Venues

  1. arXiv (Cornell University)

    Repository · 3 publications · 81 citations

    3

    h-index

  2. Geophysical Research Letters

    2 publications · 70 citations

    2

    h-index

  3. Applied Intelligence

    1 publications · 20 citations

    1

    h-index

  4. 1

    h-index

  5. Atmospheric Research

    1 publications · 179 citations

    1

    h-index

1–5