All Models are Wrong, but Many are Useful: Learning a Variable's Importance by Studying an Entire Class of Prediction Models Simultaneously

DOI10.48550/arXiv.1801.01489zbMath1436.62019arXiv1801.01489MaRDI QIDQ97217

Aaron Fisher, Cynthia Rudin, Francesca Dominici

Publication date: 4 January 2018

Full work available at URL: https://arxiv.org/abs/1801.01489

zbMATH Keywords

U-statistics; transparency; interpretable models; permutation importance; conditional variable importance; Rashomon

Mathematics Subject Classification ID

62M20: Inference from stochastic processes and prediction

62H30: Classification and discrimination; cluster analysis (statistical aspects)

62P25: Applications of statistics to social sciences

62A01: Foundations and philosophical topics in statistics

62G30: Order statistics; empirical distribution functions

Related Items

iml

Uses Software

• QCQP