Cargando…
Building more accurate decision trees with the additive tree
The expansion of machine learning to high-stakes application domains such as medicine, finance, and criminal justice, where making informed decisions requires clear understanding of the model, has increased the interest in interpretable machine learning. The widely used Classification and Regression...
Autores principales: | , , , , , , , , , |
---|---|
Formato: | Online Artículo Texto |
Lenguaje: | English |
Publicado: |
National Academy of Sciences
2019
|
Materias: | |
Acceso en línea: | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6778203/ https://www.ncbi.nlm.nih.gov/pubmed/31527280 http://dx.doi.org/10.1073/pnas.1816748116 |
_version_ | 1783456730399637504 |
---|---|
author | Luna, José Marcio Gennatas, Efstathios D. Ungar, Lyle H. Eaton, Eric Diffenderfer, Eric S. Jensen, Shane T. Simone, Charles B. Friedman, Jerome H. Solberg, Timothy D. Valdes, Gilmer |
author_facet | Luna, José Marcio Gennatas, Efstathios D. Ungar, Lyle H. Eaton, Eric Diffenderfer, Eric S. Jensen, Shane T. Simone, Charles B. Friedman, Jerome H. Solberg, Timothy D. Valdes, Gilmer |
author_sort | Luna, José Marcio |
collection | PubMed |
description | The expansion of machine learning to high-stakes application domains such as medicine, finance, and criminal justice, where making informed decisions requires clear understanding of the model, has increased the interest in interpretable machine learning. The widely used Classification and Regression Trees (CART) have played a major role in health sciences, due to their simple and intuitive explanation of predictions. Ensemble methods like gradient boosting can improve the accuracy of decision trees, but at the expense of the interpretability of the generated model. Additive models, such as those produced by gradient boosting, and full interaction models, such as CART, have been investigated largely in isolation. We show that these models exist along a spectrum, revealing previously unseen connections between these approaches. This paper introduces a rigorous formalization for the additive tree, an empirically validated learning technique for creating a single decision tree, and shows that this method can produce models equivalent to CART or gradient boosted stumps at the extremes by varying a single parameter. Although the additive tree is designed primarily to provide both the model interpretability and predictive performance needed for high-stakes applications like medicine, it also can produce decision trees represented by hybrid models between CART and boosted stumps that can outperform either of these approaches. |
format | Online Article Text |
id | pubmed-6778203 |
institution | National Center for Biotechnology Information |
language | English |
publishDate | 2019 |
publisher | National Academy of Sciences |
record_format | MEDLINE/PubMed |
spelling | pubmed-67782032019-10-09 Building more accurate decision trees with the additive tree Luna, José Marcio Gennatas, Efstathios D. Ungar, Lyle H. Eaton, Eric Diffenderfer, Eric S. Jensen, Shane T. Simone, Charles B. Friedman, Jerome H. Solberg, Timothy D. Valdes, Gilmer Proc Natl Acad Sci U S A Physical Sciences The expansion of machine learning to high-stakes application domains such as medicine, finance, and criminal justice, where making informed decisions requires clear understanding of the model, has increased the interest in interpretable machine learning. The widely used Classification and Regression Trees (CART) have played a major role in health sciences, due to their simple and intuitive explanation of predictions. Ensemble methods like gradient boosting can improve the accuracy of decision trees, but at the expense of the interpretability of the generated model. Additive models, such as those produced by gradient boosting, and full interaction models, such as CART, have been investigated largely in isolation. We show that these models exist along a spectrum, revealing previously unseen connections between these approaches. This paper introduces a rigorous formalization for the additive tree, an empirically validated learning technique for creating a single decision tree, and shows that this method can produce models equivalent to CART or gradient boosted stumps at the extremes by varying a single parameter. Although the additive tree is designed primarily to provide both the model interpretability and predictive performance needed for high-stakes applications like medicine, it also can produce decision trees represented by hybrid models between CART and boosted stumps that can outperform either of these approaches. National Academy of Sciences 2019-10-01 2019-09-16 /pmc/articles/PMC6778203/ /pubmed/31527280 http://dx.doi.org/10.1073/pnas.1816748116 Text en Copyright © 2019 the Author(s). Published by PNAS. https://creativecommons.org/licenses/by-nc-nd/4.0/ https://creativecommons.org/licenses/by-nc-nd/4.0/This open access article is distributed under Creative Commons Attribution-NonCommercial-NoDerivatives License 4.0 (CC BY-NC-ND) (https://creativecommons.org/licenses/by-nc-nd/4.0/) . |
spellingShingle | Physical Sciences Luna, José Marcio Gennatas, Efstathios D. Ungar, Lyle H. Eaton, Eric Diffenderfer, Eric S. Jensen, Shane T. Simone, Charles B. Friedman, Jerome H. Solberg, Timothy D. Valdes, Gilmer Building more accurate decision trees with the additive tree |
title | Building more accurate decision trees with the additive tree |
title_full | Building more accurate decision trees with the additive tree |
title_fullStr | Building more accurate decision trees with the additive tree |
title_full_unstemmed | Building more accurate decision trees with the additive tree |
title_short | Building more accurate decision trees with the additive tree |
title_sort | building more accurate decision trees with the additive tree |
topic | Physical Sciences |
url | https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6778203/ https://www.ncbi.nlm.nih.gov/pubmed/31527280 http://dx.doi.org/10.1073/pnas.1816748116 |
work_keys_str_mv | AT lunajosemarcio buildingmoreaccuratedecisiontreeswiththeadditivetree AT gennatasefstathiosd buildingmoreaccuratedecisiontreeswiththeadditivetree AT ungarlyleh buildingmoreaccuratedecisiontreeswiththeadditivetree AT eatoneric buildingmoreaccuratedecisiontreeswiththeadditivetree AT diffenderfererics buildingmoreaccuratedecisiontreeswiththeadditivetree AT jensenshanet buildingmoreaccuratedecisiontreeswiththeadditivetree AT simonecharlesb buildingmoreaccuratedecisiontreeswiththeadditivetree AT friedmanjeromeh buildingmoreaccuratedecisiontreeswiththeadditivetree AT solbergtimothyd buildingmoreaccuratedecisiontreeswiththeadditivetree AT valdesgilmer buildingmoreaccuratedecisiontreeswiththeadditivetree |