ASWBoost: Classification algorithm for noisy and imbalanced data based on parametric exponential loss
Fei Meng, Mei Yan, Hang Liu, Wan Xu, Hao Li · 2026
Read ASWBoost: Classification algorithm for noisy and imbalanced data based on parametric exponential loss by Fei Meng, Mei Yan, Hang Liu, Wan Xu, Hao Li (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/plos-pone-0356889.txt.
Description
Fei Meng and colleagues present ASWBoost, a strengthened version of the classic AdaBoost classification algorithm, aimed at messy real-world data where examples are noisy and one class vastly outnumbers the other. Their 2026 paper modifies the penalty function that decides how hard the algorithm pushes on hard examples, adding a tunable parameter that softens the pull of likely mislabeled outliers.
New update rules follow from that change, and the authors prove bounds on training error and analyze how the parameter shapes convergence. Across synthetic and public datasets, ASWBoost beat AdaBoost and several other boosting variants on balanced accuracy and related measures, matching a leading commercial-grade method overall. Fraud detection and medical screening are typical tasks where such stubborn, lopsided data is the norm.
Catalog record
| CREATOR | Fei Meng, Mei Yan, Hang Liu, Wan Xu, Hao Li |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 48439 |
| LICENSE | CC_BY |
| RIGHTS | CC BY |
| SOURCE | PLOS |