Title :
Signed Least Mean Kurtosis-Based Adaptive Line Enhancer
Author :
Ling, Ji-Cheng ; He, Long-Qing ; Guo, Ye-Cai
Author_Institution :
Nanjing Xiaozhuang Coll.
Abstract :
Based on the aim of the characteristic of error kurtosis and signed-error, a novel algorithm of sign least mean kurtosis based adaptive line enhancer (SLMKBALE) is proposed. Simulation results have shown that the computational load of the proposed SLMKBALE algorithm is much lower than that of the LMKBALE (least mean kurtosis based adaptive line enhancer) and as many as that of LMSBALE (least mean square based adaptive line enhancer), and the SLMKBALE algorithm has better ability to hand non-Gaussian and enhancing signal spectrum in comparison with the LMSBALE, SLMSBALE (signed LMSBALE), LMFBALE (least mean fourth based adaptive line enhancer) and LMKBALE algorithm and that the mean square error (MSE) of the proposed algorithm is the lowest in all algorithms when the MSE converges. Therefore, the SLMKBALE algorithm is useful and reliable
Keywords :
Gaussian processes; adaptive filters; adaptive signal processing; convergence; mean square error methods; LMKBALE algorithm; MSE algorithm; SLMKBALE algorithm; adaptive filtering; mean square error; nonGaussian process; signal spectrum; signed least mean kurtosis based adaptive line enhancer; Computational modeling; Convergence; Cost function; Cybernetics; Educational institutions; Equations; Gaussian noise; Helium; Least squares approximation; Line enhancers; Machine learning; Machine learning algorithms; Quantum computing; Computational load; Convergence; Error kurtosis; Sign least mean kurtosis (SLMK);
Conference_Titel :
Machine Learning and Cybernetics, 2006 International Conference on
Conference_Location :
Dalian, China
Print_ISBN :
1-4244-0061-9
DOI :
10.1109/ICMLC.2006.258968
