A Rule-Based Classification System Enhanced by Multi-Objective Genetic Algorithm
A Rule-Based Classification System Enhanced by Multi-Objective Genetic Algorithm
复制标题
DOI：
10.2991/jrnal.2017.4.3.13
复制
发表时间：
2017-01
期刊：
影响因子：
--
通讯作者：
S. Mabu;Kenzoh Azakami;M. Obayashi;T. Kuremoto
中科院分区：
文献类型：
--
作者：
S. Mabu;Kenzoh Azakami;M. Obayashi;T. Kuremoto
近年来，数据挖掘技术已经被开发用于从大数据中提取规则。然而，也有一些问题需要考虑，例如，它是很难判断哪些规则是重要的，哪些是不重要的;即使在简单的分类问题与少量的类，各种子模式要考虑潜在地存在于每个类。针对上述问题，提出了一种基于多目标遗传算法的规则聚类算法。
Recent years, data mining techniques have been developed for extracting rules from big data. However, there are some problems to be considered, for example, it is difficult to judge which rules are important and which are not important; and even in simple classification problems with the small number of classes, a various sub-patterns to be considered potentially exist in each class. To solve the above problems, a rule clustering algorithm using multiobjective genetic algorithm is proposed.