arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1606.08698cs.LGstat.APstat.ML
收藏

恢复数据移动:一个简单的二进制和多类不平衡数据的插件装袋器

Reviving Threshold-Moving: a Simple Plug-in Bagging Ensemble for Binary and Multiclass Imbalanced Data

KU Leuven(鲁汶大学);MIT Sloan Neuroeconomics Lab(麻省理工学院斯隆神经经济学实验室);Massachusetts Institute of Technology(麻省理工学院)

Guillem Collell, Drazen Prelec, Kaustubh Patil

摘要:

类不平衡是数据采掘方法应用中的一个主要障碍。处理它的一个常见做法是创建从重新采样的平衡数据中学习的分类器集合。例如,袋装决策树与随机欠采样(RUS)或合成少数过采样技术(SMOTE)相结合。然而,大多数重新排序方法需要对不同类的示例进行不对称的更改,这反过来又会在模型中引入自己的偏差。此外,这些方法需要在学习之前先验地指定性能度量。另一种方法是使用所谓的阈值移动方法,即后验改变模型的决策阈值以抵消不平衡,从而具有适应感兴趣的性能度量的潜力。令人惊讶的是,很少有人注意到结合装袋集成与阈值移动的潜力。在本文中,我们提出了概率阈值装袋(PT装袋),一个通用的插件方法,填补了这一空白。与通常的重新平衡实践相反,我们的方法保留了数据的自然类分布,从而得到校准良好的后验概率。我们还扩展了所提出的方法来处理多类数据。该方法在二进制和多类基准数据集上进行了验证。我们进行分析,提供见解所提出的方法。

英文摘要:

Class imbalance presents a major hurdle in the application of data mining methods. A common practice to deal with it is to create ensembles of classifiers that learn from resampled balanced data. For example, bagged decision trees combined with random undersampling (RUS) or the synthetic minority oversampling technique (SMOTE). However, most of the resampling methods entail asymmetric changes to the examples of different classes, which in turn can introduce its own biases in the model. Furthermore, those methods require a performance measure to be specified a priori before learning. An alternative is to use a so-called threshold-moving method that a posteriori changes the decision threshold of a model to counteract the imbalance, thus has a potential to adapt to the performance measure of interest. Surprisingly, little attention has been paid to the potential of combining bagging ensemble with threshold-moving. In this paper, we present probability thresholding bagging (PT-bagging), a versatile plug-in method that fills this gap. Contrary to usual rebalancing practice, our method preserves the natural class distribution of the data resulting in well calibrated posterior probabilities. We also extend the proposed method to handle multiclass data. The method is validated on binary and multiclass benchmark data sets. We perform analyses that provide insights into the proposed method.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑