二值分类中的假阴性或假阳性的最小化
Emphasis on the Minimization of False Negatives or False Positives in Binary Classification
摘要:
随着人类开始在当前产品中实施更多的机器学习,最小化二进制分类中的特定情况(如假阴性或假阳性)变得越来越重要。虽然有一些方法可以减少特定情况的偏差,但这些方法不是很有效,因此它们在模型中的使用很少。为此,引入了一种新方法来减少假阴性或假阳性,而不会大幅改变模型的整体性能或F1得分。这种方法涉及在预训练模型之后小心地改变输入的真实值。介绍了这种方法在各种数据集上应用的结果,有些数据集比其他数据集更复杂。通过在这些数据集上对多个模型架构进行实验,找到了最佳模型。在所有模型中,召回率或准确率都有所提高,分别最小化了假阴性或假阳性,而F1分数没有大幅下降。
英文摘要:
The minimization of specific cases in binary classification, such as false negatives or false positives, grows increasingly important as humans begin to implement more machine learning into current products. While there are a few methods to put a bias towards the reduction of specific cases, these methods aren't very effective, hence their minimal use in models. To this end, a new method is introduced to reduce the False Negatives or False positives without drastically changing the overall performance or F1 score of the model. This method involving the careful change to the real value of the input after pre-training the model. Presenting the results of this method being applied on various datasets, some being more complex than others. Through experimentation on multiple model architectures on these datasets, the best model was found. In all the models, an increase in the recall or precision, minimization of False Negatives or False Positives respectively, was shown without a large drop in F1 score.
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。