| 本文已被：浏览 5540次 下载 8027次 |
码上扫一扫！ |
|
|
| 海量信息融合方法及其在状态评价中的应用 |
|
李嘉菲1,2, 周斌1,2, 刘大有1,2, 胡亮1,2, 王峰1,2
|
|
1.吉林大学 计算机科学与技术学院, 吉林 长春 130012;2.符号计算与知识工程教育部重点实验室(吉林大学), 吉林 长春 130012
|
|
| 摘要: |
| 针对证据理论无法有效处理海量信息融合的不足,提出一种结合聚类和凸函数证据理论的海量信息融合方法,旨在解决状态评价等普遍而重要的应用问题.该方法首先基于聚类算法BIRCH对采集的海量信息进行预处理,形成多个簇;然后,针对状态评估类问题所用数据大多为数值数据和序数数据这一特点,计算每个簇的质心,并将其作为该簇的代表信息,基于广义三角模糊隶属函数对每个质心信息进行基本概率指派形成证据;最后,基于凸函数证据理论完成各证据的组合,从而完成了海量信息的融合.仿真实验结果表明:该方法既高效又合理地融合了海量信息,为海量信息融合技术的发展提供了一条探索途径. |
| 关键词: 证据理论 聚类 信息融合 海量信息 状态评价 |
| DOI：10.13328/j.cnki.jos.004632 |
| 分类号: |
| 基金项目:国家自然科学基金(61133011, 61170092, 60973088, 61202308); 国家高技术研究发展计划(863)(2011AA010101); 吉林省重点科技攻关项目(20130206046GX) |
|
| Massive Information Fusion Algorithm and Its Application in Status Evaluation |
|
LI Jia-Fei1,2, ZHOU Bin1,2, LIU Da-You1,2, HU Liang1,2, WANG Feng1,2
|
|
1.College of Computer Science and Technology, Jilin University, Changchun 130012, China;2.Key Laboratory of Symbolic Computation and Knowledge Engineering of Ministry of Education (Jilin University), Changchun 130012, China
|
| Abstract: |
| To solve the problem that the evidence theory can’t efficiently deal with the fusion of massive information, a new method combining clustering and the convex evidence theory is put forward. The method aims to solve the common and important application problems of the status evaluation. First, the famous clustering algorithm BIRCH is performed to pre-process the data, generating multiple clusters. Second, the centroid of each cluster is calculated as the representation of the cluster pertaining to the fact that most data used for status evaluation have numeric attribute or ordinal attribute. Then, to form the evidence provided by the information in each cluster, the centroid information is given a basic probability assignment value based on the generalized triangular fuzzy membership function. Finally, evidences are combined according to the combination rule of the convex evidence theory. As a result, the massive information fusion is achieved. The results of simulation experiment show that the presented method can efficiently and reasonably perform the massive information fusion, providing a new way to improve the massive information fusion techniques. |
| Key words: evidence theory clustering information fusion massive information status evaluation |
