计算机应用 ›› 2016, Vol. 36 ›› Issue (11): 2945-2949.DOI: 10.11772/j.issn.1001-9081.2016.11.2945

• 第十六届中国粗糙集与软计算联合学术会议(CRSSC 2016)论文 •    下一篇

基于压缩理论的区间概念格参数优化模型

李明霞1,2, 刘保相1,2, 张春英1,2   

  1. 1. 华北理工大学 理学院, 河北 唐山 063009;
    2. 河北省数据科学与应用重点实验室, 河北 唐山 063009
  • 收稿日期:2016-06-20 修回日期:2016-06-27 出版日期:2016-11-10 发布日期:2016-11-12
  • 通讯作者: 刘保相
  • 作者简介:李明霞(1992-),女,河北保定人,硕士研究生,主要研究方向:粗糙集、区间概念格;刘保相(1957-),男,河北衡水人,教授,主要研究方向:模糊控制、概念格、数据挖掘;张春英(1969-),女,河北唐山人,教授,博士,CCF会员,主要研究方向:概念格、人工智能、多关系数据挖掘。
  • 基金资助:
    国家自然科学基金资助项目(61370168,61472340);河北省自然科学基金资助项目(F2016209344);华北理工大学青年科学研究基金资助项目(Z201517)。

Parameter optimization model of interval concept lattice based on compression theory

LI Mingxia1,2, LIU Baoxiang1,2, ZHANG Chunying1,2   

  1. 1. College of Science, North China University of Science and Technology, Tangshan Hebei 063009, China;
    2. Key Laboratory of Data Science and Application of Hebei Province, Tangshan Hebei 063009, China
  • Received:2016-06-20 Revised:2016-06-27 Online:2016-11-10 Published:2016-11-12
  • Supported by:
    This work is partially supported by the National Natural Science Foundation of China (61370168, 61472340), the Hebei Province Natural Science Foundation (F2016209344), the Youth Science Fund of North China University of Science and Technology (Z201517).

摘要: 在由形式背景构建区间概念格之前,首先要确定区间参数[αβ],区间参数的选取影响着概念外延、格结构以及提取的关联规则数量和精度。为了获取区间概念格的压缩度达到最大时的[αβ],首先,提出了基于形式背景的二元关系对的相似度和二元关系上的覆盖近邻空间的定义,得到二元关系对的相似矩阵,并根据γ相似类求得的覆盖来计算二元关系对的近邻;其次,给出基于参数变化的概念集合更新算法,在非重建的基础上得到各区间参数下概念集合,并结合各区间参数下二元关系对的近邻空间,进一步构建基于压缩理论的区间概念格参数优化模型,依据压缩度的大小以及变化趋势寻找区间参数最优值;最后,通过实例验证了模型的有效性。

关键词: 区间概念格, 区间参数, 关系相似矩阵, 覆盖近邻空间, 压缩度

Abstract: Before building interval concept lattice from the formal context, the interval parameters[α, β] should be determined, which influence the concept extension, the lattice structure and the quantity and precision of extracted association rules. In order to obtain α and β with the biggest compression degree of interval concept lattice, firstly the definition of the similarity of binary relation pairs and covering-neighborhood-space from formal context were proposed, the similarity matrix of binary relation pairs was obtained, and the neighborhood of binary relation pairs was calculated by the covering which was obtained by similar class of γ. Secondly, update algorithm of concept sets based on change of parameters was raised, where concept sets were got on the basis of the non-reconstruction. Combining with covering-neighborhood of binary relation pairs on changing interval parameters, further the model of parameter optimization of interval concept lattice could be built based on compression theory. According to the size of the compression degree and its changing trend, the optimal values of interval parameters were found. Finally, the validity of the model was demonstrated by an example.

Key words: interval concept lattice, interval parameter, relation similarity matrix, covering-neighborhood-space, compression degree

中图分类号: