arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 1206.7112cs.LGcs.IRstat.ML
收藏

一种混合的距离度量学习方法

A Hybrid Method for Distance Metric Learning

Stanford University(斯坦福大学)

Yi-Hao Kao, Benjamin Van Roy, Daniel Rubin, Jiajing Xu, Jessica Faruque, Sandy Napel

摘要:

我们考虑的问题,学习的特征空间中的向量之间的距离的措施,并提出了一种混合的方法,同时学习分配给对的向量和类标签分配给各个向量的相似性评级。我们的方法是基于一个生成模型,其中类标签可以提供的信息,是不是在特征向量编码,但还涉及到感知对象之间的相似性。实验与合成数据以及一个真正的医学图像检索问题表明,利用类标签,通过使用我们的方法,提高检索性能显着。

英文摘要:

We consider the problem of learning a measure of distance among vectors in a feature space and propose a hybrid method that simultaneously learns from similarity ratings assigned to pairs of vectors and class labels assigned to individual vectors. Our method is based on a generative model in which class labels can provide information that is not encoded in feature vectors but yet relates to perceived similarity between objects. Experiments with synthetic data as well as a real medical image retrieval problem demonstrate that leveraging class labels through use of our method improves retrieval performance significantly.

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。

↑