Quantised kernel least mean square with desired signal smoothing

Xiguang Xu, Badong Chen, Hua Qu, Xiaohan Yang, Jihong Zhao
2015 Electronics Letters  
Quantized kernel least mean square (QKLMS) is a simple yet efficient online learning algorithm, which reduces the computational cost significantly by quantizing the input space to constrain the growth of network size. The QKLMS considers only the input space compression and assumes that the desired outputs of the quantized data are equal to those of the closest centres. In many cases, however, the outputs in a neighbourhood may have big differences, especially when underlying system is
more » ... system is disturbed by impulsive noises. Such fluctuation in desired outputs may seriously deteriorate the learning performance. To address this issue, we propose in this work a simple online method to smooth the desired signal within a neighbourhood corresponding to a quantization region. The resulting algorithm is referred to as the QKLMS with desired signal smoothing (or briefly, QKLMS-S). The desirable performance of the new algorithm is confirmed by Monte Carlo simulations.
doi:10.1049/el.2015.1757 fatcat:guh3ih3yqzg7fmvni5ljvo532a