Adapting Supervised Classification Algorithms to Arbitrary Weak Label Scenarios
Adapting Supervised Classification Algorithms to Arbitrary Weak Label Scenarios
复制标题
使监督分类算法适应任意弱标签场景
DOI：
--
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
Jesús Cid
中科院分区：
文献类型：
--
作者：
Miquel Perello;Raúl Santos;Jesús Cid
在许多现实世界的问题中，标签通常很弱，这意味着每个实例都被标记为属于多个候选类别之一，最多其中一个是真实的。最近的理论贡献表明，可以分别通过传统正确或分类校准损失的线性变换来构造弱标记分类场景的正确损失或分类校准损失。然而，如何将这些理论成果转化为实践尚未得到探索。本文讨论了算法设计和该方法的潜在优势，分析了实际设置中出现的一致性和凸性问题，并评估了不同类型弱标签下此类转换的行为。
In many real-world problems, labels are often weak, meaning that each instance is labelled as belonging to one of several candidate categories, at most one of them being true. Recent theoretical contributions have shown that it is possible to construct proper losses or classification calibrated losses for weakly labelled classification scenarios by means of a linear transformation of conventional proper or classification calibrated losses, respectively. However, how to translate these theoretical results into practice has not been explored yet. This paper discusses both the algorithmic design and the potential advantages of this approach, analyzing consistency and convexity issues arising in practical settings, and evaluating the behavior of such transformations under different types of weak labels.