基于计算几何非线性可视化分类器设计.pdfVIP

1 Vol .39 No. 1 2011 1 ACTA ELECTRONICA SINICA Jan. 2011 1, 2 2 张 涛 , 洪文学 ( 1. , 066004; 2. , 066004) : . . , , , , . , , , . UCI , , . : ; ; ; ; : TP391 : A : ( 2011) A Nonlinear Visual Classifier Based on Computational Geometry ZHANG Tao1, 2,HONG Wen ue2 ( 1. College of Inf ormation Engineering; Yanshan University ; Qinhuangdao; Hebei 066004; China; 2. College of Electrical Engineering, Yanshan University , Qinhuangdao, Hebei 066004; China) Abstract : One of the basic issues in pattern recognition is to calculate the boundary between different categories. In this pa per, we propose a novel method for that based on computational geometry named active e pansion. At first, we quantize the descrip tion space. And then term the set as base and nonbase points according their distribution, by active e panding for base points, any point in the whole space could e press the category information and the boundary is obtained. Using this method, we design the scat ter classifier which incorporates the active e pansion with combining feature attribute of scatter plot, that mapping the data from low dimension to high dimension and conforming a visual combing classifier. The e periments against UCI datasets show that perfor mance of the novel classifier has been equivalent to the popular classifiers, and outweigh in some dataset. Key words: computational geometry; active e pansion; boundary; visualization; combining classifier . , k 1 ( kNearest Neighborhood, kNN) ,