arXiv 2006.122932020-06cs.LGmath.OCstat.ML

DeepTopPush：简单且可扩展的方法，可实现最高精度

DeepTopPush: Simple and Scalable Method for Accuracy at the Top

Czech Technical University in Prague（布拉格捷克理工大学）；Czech Academy of Sciences（捷克科学院）

Václav Mácha, Lukáš Adam, Václav Šmídl

摘要：

顶部准确度是一类特殊的二元分类问题，其性能仅在少量相关（顶部）样本上进行评估。应用包括信息检索系统或带有手动（昂贵）后处理的过程。这导致最小化阈值以上的不相关样本的数量。我们考虑任意（深度）网络形式的分类器，并提出了一种新的方法DeepTopPush，用于最小化顶部的损失函数。由于阈值取决于所有样本，因此问题是不可分解的。我们修改了随机梯度下降，以端到端的训练方式处理不可分解性，并提出了一种方法，仅从当前minibatch和一个延迟值的值来估计阈值。我们展示了DeepTopPush在视觉识别数据集和两个真实世界应用程序上的出色性能。第一个选择了少量的分子进行进一步的药物测试。第二个使用真实的恶意软件数据，我们检测到46\%的恶意软件，误报率极低，为$10^{-5}$。

英文摘要：

Accuracy at the top is a special class of binary classification problems where the performance is evaluated only on a small number of relevant (top) samples. Applications include information retrieval systems or processes with manual (expensive) postprocessing. This leads to minimizing the number of irrelevant samples above a threshold. We consider classifiers in the form of an arbitrary (deep) network and propose a new method DeepTopPush for minimizing the loss function at the top. Since the threshold depends on all samples, the problem is non-decomposable. We modify the stochastic gradient descent to handle the non-decomposability in an end-to-end training manner and propose a way to estimate the threshold only from values on the current minibatch and one delayed value. We demonstrate the excellent performance of DeepTopPush on visual recognition datasets and two real-world applications. The first one selects a small number of molecules for further drug testing. The second one uses real malware data, where we detected 46\% malware at an extremely low false alarm rate of $10^{-5}$.

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。

摘要：

英文摘要：

相关论文