Learning from Imbalanced Data Sets with Weighted Cross-Entropy Function
Learning from Imbalanced Data Sets with Weighted Cross-Entropy Function
复制标题
DOI：
10.1007/s11063-018-09977-1
复制
发表时间：
2019-10-01
影响因子：
3.1
通讯作者：
Braga, Antonio Padua
中科院分区：
文献类型：
--
作者：
Aurelio, Yuri Sousa;de Almeida, Gustavo Matheus;Braga, Antonio Padua
本文提出了一种新的方法来处理不平衡数据集的神经网络问题，将先验概率的成本敏感的交叉熵误差函数。使用不同的指标，即G均值、ROC曲线下面积（AUC）、调整后的G均值、准确度、真阳性率、真阴性率和F1评分，对几个经典基准进行了性能评估测试。所得到的结果进行了比较，著名的算法，并显示了所提出的方法，这导致在不同的不平衡情况下，以及平衡的分类器的有效性和鲁棒性。
This paper presents a novel approach to deal with the imbalanced data set problem in neural networks by incorporating prior probabilities into a cost-sensitive cross-entropy error function. Several classical benchmarks were tested for performance evaluation using different metrics, namely G-Mean, area under the ROC curve (AUC), adjusted G-Mean, Accuracy, True Positive Rate, True Negative Rate and F1-score. The obtained results were compared to well-known algorithms and showed the effectiveness and robustness of the proposed approach, which results in well-balanced classifiers given different imbalance scenarios.