DocumentCode :
2509952
Title :
Arithmetic precision of the Generalized Hebbian Algorithm for hardware implementation
Author :
Kojima, Kaai ; Okuyama, Yuichi ; Kuroda, Kenichi
Author_Institution :
Grad. Sch. of Comput. Sci. & Eng., Univ. of Aizu, Fukushima
fYear :
2008
fDate :
8-11 July 2008
Firstpage :
886
Lastpage :
890
Abstract :
The principal component analysis (PCA) is a data mining methodology to express multivariate data comprehensively. The PCA reduces the dimension of data set, but its computational complexity easily gets large depending on the input factors. In this paper, we evaluate calculation accuracy of the PCA for hardware implementation. As a PCA learning algorithm, the generalized Hebbian algorithm (GHA) is adopted under the assumption of targeting field programmable gate arrays (FPGAs). With the aim of verification of the errors and required accuracy to reduce necessary hardware resources. The GHA is implemented by software in C language using input graphical images. The relationship between the three parameters, the number of principal components, mantissa bit width, and the errors, was found by comparing the output principal component images with the originals. This result will be applied to the implementation of circuit on hardware.
Keywords :
C language; Hebbian learning; circuit CAD; field programmable gate arrays; logic design; principal component analysis; C language; arithmetic precision; computational complexity; field programmable gate array; generalized Hebbian algorithm; graphical image; hardware implementation; learning algorithm; mantissa bit width; principal component analysis; Arithmetic; Circuits; Computational complexity; Covariance matrix; Data mining; Eigenvalues and eigenfunctions; Field programmable gate arrays; Hardware; Matrix decomposition; Principal component analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer and Information Technology, 2008. CIT 2008. 8th IEEE International Conference on
Conference_Location :
Sydney, NSW
Print_ISBN :
978-1-4244-2357-6
Electronic_ISBN :
978-1-4244-2358-3
Type :
conf
DOI :
10.1109/CIT.2008.4594791
Filename :
4594791
Link To Document :
بازگشت