A Statistical Approach for Learning from Order Examples by Linear Models
A Statistical Approach for Learning from Order Examples by Linear Models
复制标题
通过线性模型从订单示例中学习的统计方法
DOI：
10.11517/pjsai.jsai02.0.181.0
复制
发表时间：
2002
期刊：
影响因子：
--
通讯作者：
Toshihiro Kamishima
中科院分区：
文献类型：
--
作者：
S. Akaho;Toshihiro Kamishima
“从有序例子中学习”是一个框架，在这个框架中，系统通过使用部分有序的项目来获得用户对整个项目集的偏好。我们提出了一种基于统计线性模型的学习算法。我们还研究了算法对人工数据集的行为，并得出结论，该算法对噪声和滋扰属性具有鲁棒性。
‘Learning from order examples’ is a framework in which a system acquires a user’s preference for the whole set of items by using a partly ordered items. We provide a learning algorithm based on a statistical linear models. We also examine behaviors of the algorithm for artificial data sets, and conclude that the algorithm is robust against noise and nuisance attributes.