Title :
Polynomial Radial Base Function: A New Local Similarity Measure
Author :
Luo, Linkai ; Liu, Ting ; Peng, Hong ; Zhou, Qifeng
Author_Institution :
Dept. of Autom., Xiamen Univ., Xiamen
Abstract :
This paper studies the local similarity measure function. First, we propose polynomial radial base function (PRBF) as a local similarity measure. Then we compare the properties between Gaussian radial base function (GRBF) and PRBF, and show PRBF not only is similar to GRBF but also is of a faster computational rapidity. PRBF is a good substitute for GRBF when high computational rapidity is demanded.
Keywords :
polynomials; radial basis function networks; Gaussian radial base function; local similarity measure; polynomial radial base function; Computational efficiency; Interpolation; Pattern recognition; Polynomials; Semisupervised learning; Unsupervised learning; Gaussian Radial Base Function; Polynomial Radial Base Function; Similarity Measure;
Conference_Titel :
Natural Computation, 2008. ICNC '08. Fourth International Conference on
Conference_Location :
Jinan
Print_ISBN :
978-0-7695-3304-9
DOI :
10.1109/ICNC.2008.276
