Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2412.14527 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909433903185920 |
|---|---|
| author | Mak, Alex Sahoo, Shubham Pandey, Shivani Yue, Yidan Kong, Linglong |
| author_facet | Mak, Alex Sahoo, Shubham Pandey, Shivani Yue, Yidan Kong, Linglong |
| contents | Class imbalance and distributional differences in large datasets present significant challenges for classification tasks machine learning, often leading to biased models and poor predictive performance for minority classes. This work introduces two novel undersampling approaches: mutual information-based stratified simple random sampling and support points optimization. These methods prioritize representative data selection, effectively minimizing information loss. Empirical results across multiple classification tasks demonstrate that our methods outperform traditional undersampling techniques, achieving higher balanced classification accuracy. These findings highlight the potential of combining statistical concepts with machine learning to address class imbalance in practical applications. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_14527 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Statistical Undersampling with Mutual Information and Support Points Mak, Alex Sahoo, Shubham Pandey, Shivani Yue, Yidan Kong, Linglong Machine Learning Class imbalance and distributional differences in large datasets present significant challenges for classification tasks machine learning, often leading to biased models and poor predictive performance for minority classes. This work introduces two novel undersampling approaches: mutual information-based stratified simple random sampling and support points optimization. These methods prioritize representative data selection, effectively minimizing information loss. Empirical results across multiple classification tasks demonstrate that our methods outperform traditional undersampling techniques, achieving higher balanced classification accuracy. These findings highlight the potential of combining statistical concepts with machine learning to address class imbalance in practical applications. |
| title | Statistical Undersampling with Mutual Information and Support Points |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2412.14527 |