论文标题
分类中标签的概率解耦
Probabilistic Decoupling of Labels in Classification
论文作者
论文摘要
在本文中，我们开发了一种有原则的，概率的，统一的方法来进行非标准的分类任务，例如半监督，积极的，无标记的，多质的无标记和嘈杂标签的学习。我们在给定标签上训练分类器以预测标签 - 分布。然后，我们通过变异优化标签级过渡模型来推断基础类别分布。
In this paper we develop a principled, probabilistic, unified approach to non-standard classification tasks, such as semi-supervised, positive-unlabelled, multi-positive-unlabelled and noisy-label learning. We train a classifier on the given labels to predict the label-distribution. We then infer the underlying class-distributions by variationally optimizing a model of label-class transitions.
