Generalized Self-Organizing Network Algorithms for Cluster Analysis and Classifier Design

用于聚类分析和分类器设计的广义自组织网络算法

• 批准号: 9003252
• 负责人: James Bezdek
• 金额: $ 7.29万
• 依托单位: University of West Florida
• 依托单位国家: 美国
• 项目类别: Continuing grant
• 财政年份: 1990
• 资助国家: 美国
• 起止时间: 1990-09-01 至 1993-02-28
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9003252&HistoricalAwards=false
• 关键词: Generalized; Self; Organizing; Network; Algorithms

Dr. Bezdek intends to extend the generalized fuzzy c-means (FCM) clustering algorithm to include the Minkowsi norm family, suitable for detecting clusters in "boxy" data distributions. He will then elaborate the relationship between FCM and Kohonen's self-organizing feature maps, strengthing the theoretical foundations of each. These advances will be tested against standard algorithms using data from medical image analysis (radionuclide angiocardiography, EEG brain maps) and multisensor target detection.

Bezdek博士打算将广义模糊c-均值(FCM)聚类算法扩展到包括Minkowsi范数家族，该范数适用于检测“长方体”数据分布中的集群。然后，他将详细阐述FCM和Kohonen的自组织特征映射之间的关系，以加强各自的理论基础。这些进展将用来自医学图像分析(放射性核素心血管造影术、脑电脑图)和多传感器目标检测的数据与标准算法进行测试。