arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1808.02180cs.LGstat.ML
收藏

基于贝叶斯最优重标记的实例相关PU学习

Instance-Dependent PU Learning by Bayesian Optimal Relabeling

The University of Sydney(悉尼大学);Monash University(莫纳什大学)

Fengxiang He, Tongliang Liu, Geoffrey I Webb, Dacheng Tao

摘要:

当从正的和未标记的数据中学习时,一个强有力的假设是,正的观测是从$X$的分布中随机抽样的,条件是$Y = 1$,其中X代表特征,Y代表标记。现有的大多数算法都是在这种假设下进行优化设计的。然而,对于许多现实世界的应用,所观察到的正例依赖于条件概率$P(Y = 1|X)$,并且应该有偏地采样。在本文中,我们假设一个积极的例子与较高的$P(Y = 1|X)$是更有可能被标记,并提出了一个基于概率差距的PU学习算法。具体地说,通过将未标记的数据作为噪声负例,我们可以自动标记一组正例和负例,其标签与贝叶斯最优分类器分配的标签相同,并具有一致性保证。重新标记的例子有一个有偏的域,这是补救的核均值匹配技术。所提出的算法是无模型的,因此没有任何参数调整。实验结果表明,我们的方法在生成的数据集和真实世界的数据集上都能很好地工作。

英文摘要:

When learning from positive and unlabelled data, it is a strong assumption that the positive observations are randomly sampled from the distribution of $X$ conditional on $Y = 1$, where X stands for the feature and Y the label. Most existing algorithms are optimally designed under the assumption. However, for many real-world applications, the observed positive examples are dependent on the conditional probability $P(Y = 1|X)$ and should be sampled biasedly. In this paper, we assume that a positive example with a higher $P(Y = 1|X)$ is more likely to be labelled and propose a probabilistic-gap based PU learning algorithms. Specifically, by treating the unlabelled data as noisy negative examples, we could automatically label a group positive and negative examples whose labels are identical to the ones assigned by a Bayesian optimal classifier with a consistency guarantee. The relabelled examples have a biased domain, which is remedied by the kernel mean matching technique. The proposed algorithm is model-free and thus do not have any parameters to tune. Experimental results demonstrate that our method works well on both generated and real-world datasets.

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。

↑