DocumentCode :
682225
Title :
An new strategy for online evaluation of analog circuit performance based adaptive least squares support vector regression with double kernel RBF tuning
Author :
Huo Xing ; Qin Pengda ; Zhang Aihua
Author_Institution :
Coll. of Eng., Bohai Univ., Jinzhou, China
Volume :
1
fYear :
2013
fDate :
16-19 Aug. 2013
Firstpage :
120
Lastpage :
124
Abstract :
A novel strategy for online evaluation of analog circuit performance based on adaptive least squares support vector regression machine is proposed. Regarding reducing the computation, simultaneously, employing double kernel RBF to interfuse more flexibility to the kernel function online such as the bandwidths. And the design idea and constructed steps based on adaptive least square support vector regression with double kernel RBF tuning are introduced. Experiment adopted the typical circuit Sallen-Key low pass filter to prove the proposed evaluation strategy via the performance eight indexes. Simulation results show that the evaluation performance and the testing speed, especially the testing speed of the proposed is superior to that of the traditional LSSVR and ε-SVR, which is suit for applying online.
Keywords :
analogue circuits; circuit analysis computing; circuit tuning; least squares approximations; low-pass filters; regression analysis; support vector machines; adaptive least squares support vector regression; analog circuit performance; circuit Sallen-Key low pass filter; double kernel RBF tuning; online evaluation; Analog circuits; Educational institutions; Instruments; Kernel; Support vector machines; Testing; Training; Adaptive; Analog Circuit; Double Kernel; Evaluation; Least Square Support Vector Regression;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Electronic Measurement & Instruments (ICEMI), 2013 IEEE 11th International Conference on
Conference_Location :
Harbin
Print_ISBN :
978-1-4799-0757-1
Type :
conf
DOI :
10.1109/ICEMI.2013.6743068
Filename :
6743068
Link To Document :
بازگشت