arXiv cs.AI· Sanjeeb Dash, Soumyadip Ghosh, Joao Goncalves, Mark S. Squillante·· 2 天前AI 评分12
稀疏整数规划训练学习模型的分布鲁棒正则化
Distributionally Robust Regularization of Sparse Integer Programming Trained Learning Models
AI 导读
研究者将分布鲁棒优化（DRO）正则化扩展到混合整数稀疏规划，为可解释机器学习模型提供随稀疏参数变化的统计保证，避免交叉验证选择稀疏度的高昂计算成本。该方法在基于特征值规则集构建可解释二分类模型的场景中进行了演示。
正文 · 原文
Abstract:Building explainable machine learning models is crucial for human users to be able to interpret the proposed statistical relationship obtained from the training data. Mixed-integer optimization formulations are often used to train such models with explicit sparsity constraints, aiming to strike the right trade-off between sparsity and prediction accuracy. Existing methods to find the right choice of sparsity -- e.g., via cross-validation -- are computationally expensive. For convex model training formulations, recent advances in distributionally robust optimization (DRO) provide strong generalization while sidestepping this computational burden. We describe an extension of such regularization via DRO to mixed-integer sparse programs, providing statistical guarantees as a function of an associated sparsity parameter of the formulation. We illustrate the use of this approach in the case of building explainable binary classification models using sets of feature value rules.
| Subjects: | Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Optimization and Control (math.OC) |
| Cite as: | arXiv:2311.01994 [stat.ML] |
| (or arXiv:2311.01994v2 [stat.ML] for this version) | |
| https://doi.org/10.48550/arXiv.2311.01994 arXiv-issued DOI via DataCite |
Submission history
From: Mark Squillante [view email]
[v1]
Fri, 3 Nov 2023 15:45:34 UTC (55 KB)
[v2]
Thu, 15 Oct 2026 15:23:55 UTC (723 KB)
来源：arXiv cs.AI · arxiv.org