A Nonconvex Framework for Joint Sparse Sample Weighting and Parameter Estimation
Abstract
We propose a joint sparse sample weighting and parameter estimation (SWAPE) framework that jointly optimizes model parameters and sample weights to train a model while simultaneously assigning an individual weight to each training data point. Our framework accommodates several widely used loss functions for the model parameters, including convex, weakly convex, and difference-of-convex loss, and weakly convex sparsity functions to control the complexity of the sample weights. To solve the resulting bivariate problem, we propose an alternating minimization method in which the subproblems for updating the model parameters and weights are each approximated using majorize-minimize techniques. We provide several approximation methods for the loss functions considered in this paper, and show the overall algorithm's convergence. Numerical experiments on various learning tasks and data anomaly settings demonstrate the effectiveness of the proposed framework in minimizing the impact of undesirable data points compared to benchmark methods.
Related
- HARS: LDPC Bit-Flipping Decoding With Initial-Syndrome-Conditioned Parameter Mapping and Local-Reliability Weighting
- Keeping Score: Adaptive, Tuning-Free Loss Weighting for Score-Augmented Neural Ratio Estimation
- Sparse domination and the bi-parameter Hilbert transform
- Weighting Scheme and Construction in IPCW Survival Estimation
- Preconditioned Three-Term Conjugate Gradient Framework for Nonconvex Finite Minimax Problems
BibTeX
@misc{nojavan2026nonconvex,
title = {{A Nonconvex Framework for Joint Sparse Sample Weighting and Parameter Estimation}},
author = {Hassan Nojavan and Chul Moon and Miju Ahn},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.33957},
eprint = {2610.33957},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.33957},
}
RIS
TY - UNPB
TI - A Nonconvex Framework for Joint Sparse Sample Weighting and Parameter Estimation
AU - Hassan Nojavan
AU - Chul Moon
AU - Miju Ahn
PY - 2026
DA - 2026/10/25
PB - arXiv
DO - 10.48550/arxiv.2610.33957
UR - https://arxiv.org/abs/2610.33957
AB - We propose a joint sparse sample weighting and parameter estimation (SWAPE) framework that jointly optimizes model parameters and sample weights to train a model while simultaneously assigning an individual weight to each training data point. Our framework accommodates several widely used loss functions for the model parameters, including convex, weakly convex, and difference-of-convex loss, and weakly convex sparsity functions to control the complexity of the sample weights. To solve the resulting bivariate problem, we propose an alternating minimization method in which the subproblems for updating the model parameters and weights are each approximated using majorize-minimize techniques. We provide several approximation methods for the loss functions considered in this paper, and show the overall algorithm's convergence. Numerical experiments on various learning tasks and data anomaly settings demonstrate the effectiveness of the proposed framework in minimizing the impact of undesirable data points compared to benchmark methods.
ER -