Prediction Rule Reshaping

Matt Bonakdarpour, Sabyasachi Chatterjee, Rina Foygel Barber, John Lafferty
; Proceedings of the 35th International Conference on Machine Learning, PMLR 80:630-638, 2018.

Abstract

Two methods are proposed for high-dimensional shape-constrained regression and classification. These methods reshape pre-trained prediction rules to satisfy shape constraints like monotonicity and convexity. The first method can be applied to any pre-trained prediction rule, while the second method deals specifically with random forests. In both cases, efficient algorithms are developed for computing the estimators, and experiments are performed to demonstrate their performance on four datasets. We find that reshaping methods enforce shape constraints without compromising predictive accuracy.

Cite this Paper


BibTeX
@InProceedings{pmlr-v80-bonakdarpour18a, title = {Prediction Rule Reshaping}, author = {Bonakdarpour, Matt and Chatterjee, Sabyasachi and Barber, Rina Foygel and Lafferty, John}, booktitle = {Proceedings of the 35th International Conference on Machine Learning}, pages = {630--638}, year = {2018}, editor = {Jennifer Dy and Andreas Krause}, volume = {80}, series = {Proceedings of Machine Learning Research}, address = {Stockholmsmässan, Stockholm Sweden}, month = {10--15 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v80/bonakdarpour18a/bonakdarpour18a.pdf}, url = {http://proceedings.mlr.press/v80/bonakdarpour18a.html}, abstract = {Two methods are proposed for high-dimensional shape-constrained regression and classification. These methods reshape pre-trained prediction rules to satisfy shape constraints like monotonicity and convexity. The first method can be applied to any pre-trained prediction rule, while the second method deals specifically with random forests. In both cases, efficient algorithms are developed for computing the estimators, and experiments are performed to demonstrate their performance on four datasets. We find that reshaping methods enforce shape constraints without compromising predictive accuracy.} }
Endnote
%0 Conference Paper %T Prediction Rule Reshaping %A Matt Bonakdarpour %A Sabyasachi Chatterjee %A Rina Foygel Barber %A John Lafferty %B Proceedings of the 35th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2018 %E Jennifer Dy %E Andreas Krause %F pmlr-v80-bonakdarpour18a %I PMLR %J Proceedings of Machine Learning Research %P 630--638 %U http://proceedings.mlr.press %V 80 %W PMLR %X Two methods are proposed for high-dimensional shape-constrained regression and classification. These methods reshape pre-trained prediction rules to satisfy shape constraints like monotonicity and convexity. The first method can be applied to any pre-trained prediction rule, while the second method deals specifically with random forests. In both cases, efficient algorithms are developed for computing the estimators, and experiments are performed to demonstrate their performance on four datasets. We find that reshaping methods enforce shape constraints without compromising predictive accuracy.
APA
Bonakdarpour, M., Chatterjee, S., Barber, R.F. & Lafferty, J.. (2018). Prediction Rule Reshaping. Proceedings of the 35th International Conference on Machine Learning, in PMLR 80:630-638

Related Material