Permutation Feature Importance

Permutation Feature Importance is a model-agnostic interpretability metric that quantifies the relevance of individual features by measuring the degradation in predictive performance when the values of each feature are randomly shuffled, thereby breaking its association with the outcome variable.

111

Publications

7.7K

Citations

562

Authors

257

Institutions

Publications per year

2017–2026

106

Authors

562

  1. BB
    Bernd Bischl

    Ludwig-Maximilians-Universität München, Germany

    8 publications · 325 citations

    8

    h-index

  2. GC
    Giuseppe Casalicchio

    Ludwig-Maximilians-Universität München, Germany

    6 publications · 231 citations

    6

    h-index

  3. GK
    Gunnar König

    Ludwig-Maximilians-Universität München, Germany

    5 publications · 222 citations

    5

    h-index

  4. CM
    Christoph Molnar

    Ludwig-Maximilians-Universität München, Germany

    5 publications · 222 citations

    5

    h-index

  5. AM
    Andrés M. Bur

    University of Kansas Medical Center, United States

    3 publications · 173 citations

    3

    h-index

1–5 of 562

Venues

  1. Scientific Reports

    5 publications · 163 citations

    5

    h-index

  2. arXiv (Cornell University)

    Repository · 5 publications · 90 citations

    5

    h-index

  3. PLoS ONE

    4 publications · 67 citations

    4

    h-index

  4. Data Mining and Knowledge Discovery

    3 publications · 196 citations

    3

    h-index

  5. Cancers

    3 publications · 108 citations

    3

    h-index

1–5