GAVIN SMITH GAVIN.SMITH@NOTTINGHAM.AC.UK
Associate Professor
Model Class Reliance for Random Forests
Smith, Gavin; Mansilla Lobos, Roberto; Goulding, James
Authors
ROBERTO MANSILLA LOBOS Roberto.MansillaLobos@nottingham.ac.uk
Assistant Professor
JAMES GOULDING JAMES.GOULDING@NOTTINGHAM.AC.UK
Professor of Data Science
Abstract
Variable Importance (VI) has traditionally been cast as the process of estimating each variable's contribution to a predictive model's overall performance. Analysis of a single model instance, however, guarantees no insight into a variables relevance to underlying generative processes. Recent research has sought to address this concern via analysis of Rashomon sets-sets of alternative model instances that exhibit equivalent predictive performance to some reference model, but which take different functional forms. Measures such as Model Class Reliance (MCR) have been proposed, that are computed against Rashomon sets, in order to ascertain how much a variable must be relied on to make robust predictions, or whether alternatives exist. If MCR range is tight, we have no choice but to use a variable; if range is high then there exists competing, perhaps fairer models, that provide alternative explanations of the phenomena being examined. Applications are wide, from enabling construction of 'fairer' models in areas such as recidivism to health analytics and ethical marketing. Tractable estimation of MCR for non-linear models is currently restricted to Kernel Regression under squared loss [7]. In this paper we introduce a new technique that extends computation of Model Class Reliance (MCR) to Random Forest classifiers and regressors. The proposed approach addresses a number of open research questions, and in contrast to prior Kernel SVM MCR estimation, runs in linearithmic rather than polynomial time. Taking a fundamentally different approach to previous work, we provide a solution for this important model class, identifying situations where irrelevant covariates do not improve predictions.
Citation
Smith, G., Mansilla Lobos, R., & Goulding, J. (2020). Model Class Reliance for Random Forests. In Advances in Neural Information Processing Systems 33 pre-proceedings (NeurIPS 2020)
Presentation Conference Type | Edited Proceedings |
---|---|
Conference Name | 34th Conference on Neural Information Processing Systems (NeurIPS 2020) |
Start Date | Dec 7, 2020 |
End Date | Dec 12, 2020 |
Acceptance Date | Oct 1, 2020 |
Online Publication Date | Dec 12, 2020 |
Publication Date | Dec 12, 2020 |
Deposit Date | Dec 8, 2020 |
Publicly Available Date | Jan 7, 2021 |
Series Title | Advances in Neural Information Processing Systems |
Book Title | Advances in Neural Information Processing Systems 33 pre-proceedings (NeurIPS 2020) |
Public URL | https://nottingham-repository.worktribe.com/output/5127765 |
Publisher URL | https://papers.nips.cc/paper/2020/hash/fd512441a1a791770a6fa573d688bff5-Abstract.html |
Related Public URLs | https://nips.cc/ |
Files
Model Class
(1 Mb)
PDF
You might also like
Towards optimal symbolization for time series comparisons
(2013)
Presentation / Conference Contribution
A refined limit on the predictability of human mobility
(2014)
Presentation / Conference Contribution
The potential of electromyography to aid personal navigation
(2014)
Presentation / Conference Contribution
A novel symbolization technique for time-series outlier detection
(2015)
Presentation / Conference Contribution
Event series prediction via non-homogeneous Poisson process modelling
(2016)
Presentation / Conference Contribution
Downloadable Citations
About Repository@Nottingham
Administrator e-mail: discovery-access-systems@nottingham.ac.uk
This application uses the following open-source libraries:
SheetJS Community Edition
Apache License Version 2.0 (http://www.apache.org/licenses/)
PDF.js
Apache License Version 2.0 (http://www.apache.org/licenses/)
Font Awesome
SIL OFL 1.1 (http://scripts.sil.org/OFL)
MIT License (http://opensource.org/licenses/mit-license.html)
CC BY 3.0 ( http://creativecommons.org/licenses/by/3.0/)
Powered by Worktribe © 2024
Advanced Search