Controlling Constructive Induction in CIPF: An MDL Approach
Abstract
We describe the propositional learning system C i PF, which tightly couples a simple concept learner with a sophisticated constructive induction component. It is described in terms of a generic architecture for constructive induction. We focus on the problem of controlling the abundance of opportunities for constructively adding new attributes. In C i PF the so-called Minimum Description Length (MDL) principle acts as a powerful control heuristic. This is also confirmed in the experiments reported.
Cite
Text
Pfahringer. "Controlling Constructive Induction in CIPF: An MDL Approach." European Conference on Machine Learning, 1994. doi:10.1007/3-540-57868-4_62Markdown
[Pfahringer. "Controlling Constructive Induction in CIPF: An MDL Approach." European Conference on Machine Learning, 1994.](https://mlanthology.org/ecmlpkdd/1994/pfahringer1994ecml-controlling/) doi:10.1007/3-540-57868-4_62BibTeX
@inproceedings{pfahringer1994ecml-controlling,
title = {{Controlling Constructive Induction in CIPF: An MDL Approach}},
author = {Pfahringer, Bernhard},
booktitle = {European Conference on Machine Learning},
year = {1994},
pages = {242-256},
doi = {10.1007/3-540-57868-4_62},
url = {https://mlanthology.org/ecmlpkdd/1994/pfahringer1994ecml-controlling/}
}