Adaptive Feature-Space Conformal Transformation for Imbalanced-Data Learning
Abstract
When the training instances of the target class are heavily outnumbered by non-target training instances, SVMs can be ineffective in determining the class boundary. To remedy this problem, we propose an adaptive conformal transformation (ACT) algorithm. ACT considers feature-space distance and the class-imbalance ratio when it performs conformal transformation on a kernel function. Experimental results on UCI and real-world datasets show ACT to be effective in improving class prediction accuracy. ICML Proceedings of the Twentieth International Conference on Machine Learning
Cite
Text
Wu and Chang. "Adaptive Feature-Space Conformal Transformation for Imbalanced-Data Learning." International Conference on Machine Learning, 2003.Markdown
[Wu and Chang. "Adaptive Feature-Space Conformal Transformation for Imbalanced-Data Learning." International Conference on Machine Learning, 2003.](https://mlanthology.org/icml/2003/wu2003icml-adaptive/)BibTeX
@inproceedings{wu2003icml-adaptive,
title = {{Adaptive Feature-Space Conformal Transformation for Imbalanced-Data Learning}},
author = {Wu, Gang and Chang, Edward Y.},
booktitle = {International Conference on Machine Learning},
year = {2003},
pages = {816-823},
url = {https://mlanthology.org/icml/2003/wu2003icml-adaptive/}
}