KLMAT: A Kernel Least Mean Absolute Third Algorithm
For researchers in adaptive filtering and time series prediction, this is an incremental hybrid algorithm that extends existing LMAT to kernel methods with a variable step-size enhancement.
The paper develops a kernel least mean absolute third (KLMAT) algorithm for adaptive prediction, which combines kernel methods with the LMAT algorithm to achieve robustness against various noise distributions. A variable step-size version (VSS-KLMAT) is also proposed to improve convergence. Simulations show effectiveness in time series prediction, but no concrete numerical results are provided.
In this paper, a kernel least mean absolute third (KLMAT) algorithm is developed for adaptive prediction. Combining the benefits of the kernel method and the least mean absolute third (LMAT) algorithm, the proposed KLMAT algorithm performs robustly against noise with different probability densities. To further enhance the convergence rate of the KLMAT algorithm, a variable step-size version (VSS-KLMAT algorithm) is proposed based on a Lorentzian function. Moreover, the stability and convergence property of the proposed algorithms are analyzed. Simulation results in the context of time series prediction demonstrate that the effectiveness of proposed algorithms.