Investigating accuracies of rule evaluation models on randomized labeling and human evaluation
Investigating accuracies of rule evaluation models on randomized labeling and human evaluation
复制标题
DOI：
10.1109/grc.2008.4664770
复制
发表时间：
2008-10
期刊：
影响因子：
--
通讯作者：
H. Abe;S. Tsumoto
中科院分区：
文献类型：
--
作者：
H. Abe;S. Tsumoto
在数据挖掘的后处理中，利用客观规则评价指标进行规则选择是从挖掘的模式中发现有价值知识的有效方法之一。然而，指标值与专家标准之间的关系从未得到澄清。为了确定这种关系，我们开发了一种从由客观规则评价指标和规则评价标签组成的数据集中获得学习模型的方法。在本研究中，我们比较了具有随机分类分布的数据集的分类学习算法的准确性。然后，研究结果表明，在平衡随机分类分布下，有和没有人类专家标准的分类学习算法的准确率是不同的。
In datamining post-processing, rule selection using objective rule evaluation indices is one of a useful method to find out valuable knowledge from mined patterns. However, the relationship between an index value and expertspsila criteria has never been clarified. In order to determine the relationship, we have developed a method to obtain learning models from a dataset consisting of objective rule evaluation indices and evaluation labels for rules. In this study, we have compared the accuracies of classification learning algorithms for datasets with randomized class distributions. Then, the results show that accuracies of classification learning algorithms with/without criteria of human experts are different on a balanced randomized class distribution.