Title :
Target recognition of MMW radar based on FSVM
Author :
Luan, Ying-hong ; Yue-hua ; Li, Yue-hua
Author_Institution :
Inst. of Near Sensing Tech. with Millimeter Wave&Opt.-wave, Nanjing Univ. of Sci.&Tech., Nanjing
Abstract :
The choice of fuzzy membership relates to the classification results of fuzzy support vector machine (FSVM). In this paper, membership based on distance and affinity among samples is researched. Aimed at the disadvantage of membership based on affinity among samples that can not reflect the distribution of radar echo in feature space, super-ellipse is introduced to improve the membership and applied to FSVM to distinguish support vectors with outliers or noise. The results of experiments show that the fuzzy support vector machine, based on the improved affinity among samples, is more robust than others.
Keywords :
fuzzy set theory; radar target recognition; support vector machines; MMW radar; fuzzy support vector machine; membership decision; sample affinity; target recognition; Ellipsoids; Isolation technology; Machine learning; Millimeter wave radar; Millimeter wave technology; Pattern recognition; Spaceborne radar; Support vector machine classification; Support vector machines; Target recognition; FSVM; affinity; membership decision; target recognition;
Conference_Titel :
Industrial Electronics and Applications, 2008. ICIEA 2008. 3rd IEEE Conference on
Conference_Location :
Singapore
Print_ISBN :
978-1-4244-1717-9
Electronic_ISBN :
978-1-4244-1718-6
DOI :
10.1109/ICIEA.2008.4582752
