Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.01397 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911132259713024 |
|---|---|
| author | Vigl, Matthias Heinrich, Lukas |
| author_facet | Vigl, Matthias Heinrich, Lukas |
| contents | Recently, the benefit of heavily overparameterized models has been observed in machine learning tasks: models with enough capacity to easily cross the \emph{interpolation threshold} improve in generalization error compared to the classical bias-variance tradeoff regime. We demonstrate this behavior for the first time in particle physics data and explore when and where `double descent' appears and under which circumstances overparameterization results in a performance gain. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_01397 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Double Descent and Overparameterization in Particle Physics Data Vigl, Matthias Heinrich, Lukas High Energy Physics - Experiment Machine Learning High Energy Physics - Phenomenology Data Analysis, Statistics and Probability Recently, the benefit of heavily overparameterized models has been observed in machine learning tasks: models with enough capacity to easily cross the \emph{interpolation threshold} improve in generalization error compared to the classical bias-variance tradeoff regime. We demonstrate this behavior for the first time in particle physics data and explore when and where `double descent' appears and under which circumstances overparameterization results in a performance gain. |
| title | Double Descent and Overparameterization in Particle Physics Data |
| topic | High Energy Physics - Experiment Machine Learning High Energy Physics - Phenomenology Data Analysis, Statistics and Probability |
| url | https://arxiv.org/abs/2509.01397 |