Saved in:
Bibliographic Details
Main Author: Vo, Thien Nhan
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2505.03617
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911008650428416
author Vo, Thien Nhan
author_facet Vo, Thien Nhan
contents We evaluate the effectiveness of importance weighting in deep neural networks under label shift and covariate shift. On synthetic 2D data (linearly separable and moon-shaped) using logistic regression and MLPs, we observe that weighting strongly affects decision boundaries early in training but fades with prolonged optimization. On CIFAR-10 with various class imbalances, only L2 regularization (not dropout) helps preserve weighting effects. In a covariate-shift experiment, importance weighting yields no significant performance gain, highlighting challenges on complex data. Our results call into question the practical utility of importance weighting for real-world distribution shifts.
format Preprint
id arxiv_https___arxiv_org_abs_2505_03617
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Understand the Effect of Importance Weighting in Deep Learning on Dataset Shift
Vo, Thien Nhan
Machine Learning
We evaluate the effectiveness of importance weighting in deep neural networks under label shift and covariate shift. On synthetic 2D data (linearly separable and moon-shaped) using logistic regression and MLPs, we observe that weighting strongly affects decision boundaries early in training but fades with prolonged optimization. On CIFAR-10 with various class imbalances, only L2 regularization (not dropout) helps preserve weighting effects. In a covariate-shift experiment, importance weighting yields no significant performance gain, highlighting challenges on complex data. Our results call into question the practical utility of importance weighting for real-world distribution shifts.
title Understand the Effect of Importance Weighting in Deep Learning on Dataset Shift
topic Machine Learning
url https://arxiv.org/abs/2505.03617