发表机构
Warsaw University of Life Sciences（华沙生命科学大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究提出在SCAR假设违反时，结合SMOTE过采样与聚类辅助逻辑回归（含Lasso变体）解决PU分类问题，实验表明SMOTE提升性能且LassoJoint鲁棒。
AI 中文摘要
本研究针对SCAR假设被违反情况下的PU分类问题。我们研究了基于逻辑回归的方法，即聚类方法及其带有严格和非严格Lasso正则化的扩展。本研究的主要贡献是集成SMOTE技术以缓解类别不平衡，并系统评估其对所考虑算法性能的影响。首先应用SMOTE对训练数据集进行重平衡。接下来，通过2-means聚类导出清洗标签。然后在清洗后的数据上训练逻辑回归，其中识别出的正例被额外的真正例增强，其余观测被视为负例。实验评估在13个真实基准数据集和一个合成数据集上进行。作为比较，我们包括了朴素方法和Spy-EM方法。结果表明，当SCAR条件被违反时，纳入SMOTE提高了分类性能，并表明LassoJoint方法在此设置下具有中等鲁棒性。
英文摘要
This study addresses the PU classification problem under violations of the SCAR assumption. We investigate logistic regression-based approaches, namely the cluster method and its extensions with strict and non-strict Lasso regularization. The primary contribution of this work is the integration of the SMOTE technique to alleviate class imbalance and systematically assess its impact on the performance of the considered algorithms. SMOTE is first applied to rebalance the training dataset. Next, cleaning labels are derived via 2-means clustering. Logistic regression is then trained on the cleaned data, where identified positive instances are augmented with additional true positives and the remaining observations are treated as negative. The experimental evaluation is conducted on 13 real benchmark datasets and one synthetic dataset. For comparison, we include the naive approach and the Spy-EM method. The results demonstrate that incorporating SMOTE improves classification performance when the SCAR condition is violated and indicate moderate robustness of the LassoJoint method in this setting.
Comments11 pages, 1 figure, 3 tables. Supplementary materials and full reproducible R code available at: https://github.com/kapacc/mdai26
Journal refUSB Proceedings of the 23nd International Conference on Modeling Decisions for Artificial Intelligence, MDAI 2026, Vic, Catalonia, Spain 7- 9 September 2026, ISBN: 978-91-531-0241-0
