Heuristic Rule-Based Regression via Dynamic Reduction to Classification
Abstract
In this paper, we propose a novel approach for learning regression rules by transforming the regression problem into a classification problem. Unlike previous approaches to regression by classification, in our approach the discretization of the class variable is tightly integrated into the rule learning algorithm. The key idea is to dynamically define a region around the target value predicted by the rule, and considering all examples within that region as positive and all examples outside that region as negative. In this way, conventional rule learning heuristics may be used for inducing regression rules. Our results show that our heuristic algorithm outperforms approaches that use a static discretization of the target variable, and performs en par with other comparable rule-based approaches, albeit without reaching the performance of statistical approaches.
Cite
Text
Janssen and Fürnkranz. "Heuristic Rule-Based Regression via Dynamic Reduction to Classification." International Joint Conference on Artificial Intelligence, 2011. doi:10.5591/978-1-57735-516-8/IJCAI11-225Markdown
[Janssen and Fürnkranz. "Heuristic Rule-Based Regression via Dynamic Reduction to Classification." International Joint Conference on Artificial Intelligence, 2011.](https://mlanthology.org/ijcai/2011/janssen2011ijcai-heuristic/) doi:10.5591/978-1-57735-516-8/IJCAI11-225BibTeX
@inproceedings{janssen2011ijcai-heuristic,
title = {{Heuristic Rule-Based Regression via Dynamic Reduction to Classification}},
author = {Janssen, Frederik and Fürnkranz, Johannes},
booktitle = {International Joint Conference on Artificial Intelligence},
year = {2011},
pages = {1330-1335},
doi = {10.5591/978-1-57735-516-8/IJCAI11-225},
url = {https://mlanthology.org/ijcai/2011/janssen2011ijcai-heuristic/}
}