Unsupervised ranking of multi-attribute objects based on principal curves
Unsupervised ranking of multi-attribute objects based on principal curves
复制标题
基于主曲线的多属性对象无监督排序
DOI：
10.1109/icde.2016.7498407
复制
发表时间：
2014-02
影响因子：
8.9
通讯作者：
胡包钢
中科院分区：
文献类型：
--
作者：
Chun-Guo Li;Xing Mei;胡包钢
无监督排序在评估应用中面临一个关键挑战，即没有可用的基础事实。虽然PageRank及其变体在相关对象中显示了良好的解决方案，但它们仅适用于链接结构数据的排名。在这项工作中，我们专注于多属性数据的无监督排名，这在评估任务中也很常见。为了克服这一挑战，我们提出了五个基本的元规则的设计和评估的无监督排名方法：规模和平移不变性，严格的单调性，线性和非线性的兼容性，平滑性，和明确的参数大小。这些元规则被认为是无监督排名任务的高级知识。受[2]和[6]中工作的启发，我们提出了一种排名主曲线（RPC）模型，该模型学习一维流形函数来对多属性观测执行无监督排名任务。此外，RPC被建模为三次贝塞尔曲线，控制点限制在超立方体的内部，遵守所有五个元规则来推断合理的排名列表。以控制点作为模型参数，人们能够理解学习的流形，并解释和可视化排名结果。在两个不同排序应用的开放数据集上进行了RPC模型的数值实验。与现有方法相比，新模型能够给出更合理的排序结果。
Unsupervised ranking faces one critical challenge in evaluation applications, that is, no ground truth is available. While PageRank and its variants show a good solution in related objects, they are applicable only for ranking from link-structure data. In this work, we focus on unsupervised ranking from multi-attribute data which is also common in evaluation tasks. To overcome the challenge, we propose five essential meta-rules for the design and assessment of unsupervised ranking approaches: scale and translation invariance, strict monotonicity, compatibility of linearity and nonlinearity, smoothness, and explicitness of parameter size. These meta-rules are regarded as high level knowledge for unsupervised ranking tasks. Inspired by the works in [2] and [6], we propose a ranking principal curve (RPC) model, which learns a one-dimensional manifold function to perform unsupervised ranking tasks on multi-attribute observations. Furthermore, the RPC is modeled to be a cubic Bézier curve with control points restricted in the interior of a hypercube, complying with all the five meta-rules to infer a reasonable ranking list. With control points as model parameters, one is able to understand the learned manifold and to interpret and visualize the ranking results. Numerical experiments of the presented RPC model are conducted on two open datasets of different ranking applications. In comparison with the state-of-the-art approaches, the new model is able to show more reasonable ranking lists.
登录
查看更多内容
DOI：
--
发表时间：
1998-09
期刊：
--
影响因子：
--
作者：
T. Pastva
通讯作者：
T. Pastva
影响因子：
2.5
作者：
A. Vogler
通讯作者：
A. Vogler
影响因子：
3.7
作者：
T. Burr
通讯作者：
T. Burr
DOI：
10.1016/s0005-1098(98)00060-0
发表时间：
1998-09
期刊：
Autom.
影响因子：
--
作者：
Bao-Gang Hu;G. Mann;R. Gosine
通讯作者：
Bao-Gang Hu;G. Mann;R. Gosine
DOI：
10.1007/3-540-47797-7_2
发表时间：
2000-04
期刊：
--
影响因子：
--
作者：
H. Priestley
通讯作者：
H. Priestley