DMCA
The Case Against Accuracy Estimation for Comparing Induction Algorithms (1997)
Cached
Download Links
| Venue: | In Proceedings of the Fifteenth International Conference on Machine Learning |
| Citations: | 414 - 23 self |
BibTeX
@INPROCEEDINGS{Provost97thecase,
author = {Foster Provost and Tom Fawcett and Ron Kohavi},
title = {The Case Against Accuracy Estimation for Comparing Induction Algorithms},
booktitle = {In Proceedings of the Fifteenth International Conference on Machine Learning},
year = {1997},
pages = {445--453},
publisher = {Morgan Kaufmann}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
We analyze critically the use of classification accuracy to compare classifiers on natural data sets, providing a thorough investigation using ROC analysis, standard machine learning algorithms, and standard benchmark data sets. The results raise serious concerns about the use of accuracy for comparing classifiers and drawinto question the conclusions that can be drawn from such studies. In the course of the presentation, we describe and demonstrate what we believe to be the proper use of ROC analysis for comparative studies in machine learning research. We argue that this methodology is preferable both for making practical choices and for drawing scientific conclusions.
Keyphrases
comparing induction algorithm case accuracy estimation roc analysis serious concern standard machine natural data set practical choice machine learning research classification accuracy proper use scientific conclusion standard benchmark data set thorough investigation drawinto question comparative study
