We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
STOCHASTIC INFORMATION GRADIENT ALGORITHM WITH GENERALIZED GAUSSIAN DISTRIBUTION MODEL.
- Authors
CHEN, BADONG; PRINCIPE, JOSE C.; HU, JINCHUN; ZHU, YU
- Abstract
This paper presents a parameterized version of the stochastic information gradient (SIG) algorithm, in which the error distribution is modeled by generalized Gaussian density (GGD), with location, shape, and dispersion parameters. Compared with the kernel-based SIG (SIG-Kernel) algorithm, the GGD-based SIG (SIG-GGD) algorithm does not involve kernel width selection. If the error is zero-mean, the SIG-GGD algorithm will become the least mean p-power (LMP) algorithm with adaptive order and variable step-size. Due to its well matched density estimation and automatic switching capability, the proposed algorithm is favorably in line with existing algorithms.
- Subjects
STOCHASTIC information theory; ALGORITHMS; GAUSSIAN distribution; KERNEL functions; SWITCHING circuits; ENTROPY (Information theory)
- Publication
Journal of Circuits, Systems & Computers, 2012, Vol 21, Issue 1, p1250006-1
- ISSN
0218-1266
- Publication type
Article
- DOI
10.1142/S0218126612500065