arXiv：cs.LG· Daniela De Canditiis, Italia De Feis, Paola Stolfi·· 13 小时前
Robust LassoNet：用鲁棒损失函数增强神经网络特征选择
Robust LassoNet: Enhancing Feature Selection in Neural Networks via Robust Loss Functions
AI 导读
Robust LassoNet 在 LassoNet 基础上引入 Huber、Cauchy、Tukey's bisquare 和 Nonnegative Garrote 等鲁棒损失函数，以缓解原始 MSE 损失对异常值高度敏感的问题，同时保留原有优化框架。在合成与真实数据集上的实验显示，该方法在存在异常值或重尾噪声时显著提升了预测性能与特征选择准确率，而在干净数据上性能相当。
正文
Abstract:Feature selection in neural networks remains a challenging problem, particularly in the presence of noisy or contaminated data. LassoNet is a recent approach that addresses this issue by combining neural networks with hierarchical sparsity constraints, enabling simultaneous prediction and variable selection. However, its standard formulation relies on the mean squared error (MSE) loss, which is known to be highly sensitive to outliers. In this paper, we present Robust LassoNet, an extension of LassoNet that incorporates robust loss functions, such as Huber, Cauchy, Tukey's bisquare, and Nonnegative Garrote, to mitigate the effect of extreme observations. The proposed approach preserves the original optimization framework while improving stability under data contamination. Through experiments on synthetic and real datasets, we show that robust LassoNet significantly improves both predictive performance and feature selection accuracy in the presence of outliers or heavy tailed noise, while maintaining comparable performance in clean settings. These results highlight the importance of robustness in neural network based feature selection and suggest practical guidelines for choosing appropriate loss functions.
| Subjects: | Machine Learning (stat.ML); Machine Learning (cs.LG); Probability (math.PR) |
| Cite as: | arXiv:2609.38263 [stat.ML] |
| (or arXiv:2609.38263v2 [stat.ML] for this version) | |
| https://doi.org/10.48550/arXiv.2609.38263 arXiv-issued DOI via DataCite |
Submission history
From: Paola Stolfi [view email]
[v1]
Tue, 29 Sep 2026 12:49:41 UTC (9,564 KB)
[v2]
Fri, 9 Oct 2026 09:27:14 UTC (9,564 KB)
来源：arXiv：cs.LG · arxiv.org