基于可交互相关性矩阵的维度重排径向坐标可视化方法  

Interactive Dimension Reordering in RadViz with Correlation Matrix

作　　者：张志豪[1,2] 张军平[1,2] 陈德铭 卢颖[1] 袁晓如[4] 古天龙[5] ZHANG Zhihao ZHANG Junping CHAN Takming LU Ying YUAN Xiaoru GU Tianlong(School of Computer Science, Fudan University, Shanghai 200433 Shanghai Key Laboratory of Intelligent Information Processing, Fudan University, Shanghai 200433 Philips Research China, Shanghai 201102 School of Electronics Engineering and Computer Science, Peking University, Belting 100871 Guangxi Key Laboratory of Trusted Software, Guilin University of Electronic Technology, Guilin 541004)

机构地区：[1]复旦大学计算机科学技术学院,上海200433 [2]复旦大学上海市智能信息处理重点实验室,上海200433 [3]飞利浦中国研究院,上海201102 [4]北京大学信息科学技术学院,北京100871 [5]桂林电子科技大学广西可信软件重点实验室,桂林541004

出　　处：《模式识别与人工智能》2017年第7期637-645,共9页Pattern Recognition and Artificial Intelligence

基　　金：国家自然科学基金项目(No.61673118;61572146;U1501252);上海市浦江人才计划项目(No.16PJD009)资助~~

摘　　要：针对现有维度排序方法较少考虑到用户参与的问题,文中提出基于可交互相关性矩阵的维度重排径向坐标可视化方法.采用考虑到径向坐标可视化(RadViz)特性、拥有可变参数的层次聚类算法推荐初始的维度顺序,提供展示算法结果的系统树图用于指导用户交互式地调整、选取、删除维度,进行特征子集选择.实验表明,文中方法具有较好的交互性,注重用户体验,减轻RadViz中的投影点重叠问题.In the existing dimension reordering algorithms, the interactive functions allowing users to perform order navigation or reorder operations are rarely taken into account. Aiming at this problem, a navigation- guided correlation matrix is proposed for users to interactively reorder dimensions in radial visualization （RadViz）. A hierarchical clustering algorithm with configurable parameters is specially designed for RadViz to recommend the initial dimension order. The dendrogram of results is employed to help users interactively reorder, select and delete dimensions for feature subset selection. Experiments show that the proposed method is interactive, user-friendly and helpful for alleviating the overlapping problem of data projections in RadViz.

关 键 词：径向坐标可视化(RadViz) 维度重排 相关性矩阵 层次聚类 特征子集选择 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]