Latent Supervised Learning.

Latent Supervised Learning.
复制标题

DOI:
10.1080/01621459.2013.789695
复制
发表时间:
2013-07-01
影响因子:
3.7
通讯作者:
Kosorok MR
Kosorok MR
中科院分区:
数学1区
文献类型:
--
作者:
Wei S;Kosorok MR

文献摘要

参考文献

被引文献

相似文献

引入了一种新的机器学习任务,称为潜在监督学习,其目标是从连续训练标签中学习二元分类器,这些标签充当未观察到的类标签的替代品。研究了一个特定模型,其中代理变量来自均值和方差未知的二分量高斯混合,并且分量隶属度由协变量空间中的超平面确定。分离超平面和高斯混合参数的估计形成了所谓的变线分类问题。提出了一种数据驱动的超平面筛最大似然估计器,该估计器又可用于估计高斯混合的参数。估计量被证明是一致的。模拟和经验数据表明估计器具有很高的分类精度。
A new machine learning task is introduced, called latent supervised learning, where the goal is to learn a binary classifier from continuous training labels which serve as surrogates for the unobserved class labels. A specific model is investigated where the surrogate variable arises from a two-component Gaussian mixture with unknown means and variances, and the component membership is determined by a hyperplane in the covariate space. The estimation of the separating hyperplane and the Gaussian mixture parameters forms what shall be referred to as the change-line classification problem. A data-driven sieve maximum likelihood estimator for the hyperplane is proposed, which in turn can be used to estimate the parameters of the Gaussian mixture. The estimator is shown to be consistent. Simulations as well as empirical data show the estimator has high classification accuracy.
DOI: 10.1214/aos/1176345782
发表时间: 1982-01-01
影响因子: 4.5
作者:
GEMAN, S;HWANG, CR
通讯作者: HWANG, CR
DOI: 10.2307/2290563
发表时间: 1991-06-01
影响因子: 3.7
作者:
LI, KC
通讯作者: LI, KC
DOI: 10.1377/hlthaff.24.1.67
发表时间: 2005-01-01
期刊: HEALTH AFFAIRS
影响因子: 9.7
作者:
Fleming, TR
通讯作者: Fleming, TR
DOI: 10.2307/2669998
发表时间: 1999-09-01
影响因子: 3.7
作者:
Shen, XT;Shi, J;Wong, WH
通讯作者: Wong, WH
DOI: 10.1073/pnas.082099299
发表时间: 2002-05-14
影响因子: 11.1
作者:
Tibshirani, R;Hastie, T;Chu, G
通讯作者: Chu, G