Kernel Weibull M-transform Least Mean Square Algorithm

The paper introduces a Kernel Weibull M-transform Least-Mean Square (LMS) (KWMLMS) algorithm aimed to enhance filtering performance for a nonlinear system. By incorporating the Weibull M-transformation into the cost function of the Kernel LMS (KLMS) algorithm, a weighting updated rule is gotten. Simulation experiments conducted for nonlinear-system identification under impulse noise conditions demonstrate that the created KWMLMS algorithm achieves significantly faster convergence compared to existing methods, indicating its superiority in dealing with impulsive-noises.