Home > Research > Publications & Outputs > An Empirical Evaluation of the Rashomon Effect ...

Research

Electronic data

Links

Keywords

View graph of relations

An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning

Research output: Working paper › Preprint

Published

Standard

An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning. / Müller, Sebastian; Toborek, Vanessa; Beckh, Katharina et al.
Arxiv, 2023.

Research output: Working paper › Preprint

Harvard

Müller, S, Toborek, V, Beckh, K, Jakobs, M, Bauckhage, C & Welke, P 2023 'An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning' Arxiv. <https://arxiv.org/abs/2306.15786v2#>

APA

Müller, S., Toborek, V., Beckh, K., Jakobs, M., Bauckhage, C., & Welke, P. (2023). An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning. Arxiv. https://arxiv.org/abs/2306.15786v2#

Vancouver

Müller S, Toborek V, Beckh K, Jakobs M, Bauckhage C, Welke P. An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning. Arxiv. 2023 Jun 27.

Author

Müller, Sebastian ; Toborek, Vanessa ; Beckh, Katharina et al. / An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning. Arxiv, 2023.

Bibtex

@techreport{1c27cf2b4bf94d3c8432c1597dbf54bf,
title = "An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning",
abstract = " The Rashomon Effect describes the following phenomenon: for a given dataset there may exist many models with equally good performance but with different solution strategies. The Rashomon Effect has implications for Explainable Machine Learning, especially for the comparability of explanations. We provide a unified view on three different comparison scenarios and conduct a quantitative evaluation across different datasets, models, attribution methods, and metrics. We find that hyperparameter-tuning plays a role and that metric selection matters. Our results provide empirical support for previously anecdotal evidence and exhibit challenges for both scientists and practitioners. ",
keywords = "cs.LG, cs.AI",
author = "Sebastian M{\"u}ller and Vanessa Toborek and Katharina Beckh and Matthias Jakobs and Christian Bauckhage and Pascal Welke",
year = "2023",
month = jun,
day = "27",
language = "English",
publisher = "Arxiv",
type = "WorkingPaper",
institution = "Arxiv",

}

RIS

TY - UNPB

T1 - An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning

AU - Müller, Sebastian

AU - Toborek, Vanessa

AU - Beckh, Katharina

AU - Jakobs, Matthias

AU - Bauckhage, Christian

AU - Welke, Pascal

PY - 2023/6/27

Y1 - 2023/6/27

N2 - The Rashomon Effect describes the following phenomenon: for a given dataset there may exist many models with equally good performance but with different solution strategies. The Rashomon Effect has implications for Explainable Machine Learning, especially for the comparability of explanations. We provide a unified view on three different comparison scenarios and conduct a quantitative evaluation across different datasets, models, attribution methods, and metrics. We find that hyperparameter-tuning plays a role and that metric selection matters. Our results provide empirical support for previously anecdotal evidence and exhibit challenges for both scientists and practitioners.

AB - The Rashomon Effect describes the following phenomenon: for a given dataset there may exist many models with equally good performance but with different solution strategies. The Rashomon Effect has implications for Explainable Machine Learning, especially for the comparability of explanations. We provide a unified view on three different comparison scenarios and conduct a quantitative evaluation across different datasets, models, attribution methods, and metrics. We find that hyperparameter-tuning plays a role and that metric selection matters. Our results provide empirical support for previously anecdotal evidence and exhibit challenges for both scientists and practitioners.

KW - cs.LG

KW - cs.AI

M3 - Preprint

BT - An Empirical Evaluation of the Rashomon Effect in Explainable Machine Learning

PB - Arxiv

ER -