RADIAL-BASIS-FUZZY-WAVELET-NEURAL NETWORK WITH ADAPTIVE
ACTIVATION-MEMBERSHIP FUNCTION
Ye.
Bodyanskiy, O. Vynokurova, E. Yegorova
Control System
Research Laboratory,
Kharkiv National
University of Radio Electronics
av. Lenina, 14,
CSRL, Kharkiv, Ukraine
Abstract
In this paper a two-layers architecture of
computational intelligence hybrid fuzzy-wavelet-neuro-system
is considered. Hidden layer of this architecture is formed
of the wavelons with adaptive activation-membership
function, and output layer is formed of the adaptive linear
associators. The proposed adaptive learning algorithm has
both following and filtering properties, allows a real time
nonstationary nonlinearly signals processing and provides
the significant improvement of approximating properties what
is shown in the results of experimental simulation.
Keywords:
Computational intelligence, hybrid neural networks,
wavelet membership function, forecasting, emulation,
fuzzy-wavelet-neural network.
( P1120827003,
971 KB)
Biographies:
 |
Prof. Yevgeniy Bodyanskiy, Scientific Head of the
Control Systems Research Laboratory in Kharkiv
National University of Radio Electronics (KNURE)
Ukraine, Senior Member of IEEE, a member of three
scientific and five editorial boards. He received
his Master degree in Automatic and Remote Control in
1971 in KNURE, Ph.D. in Technical Cybernetics and
Information Theory in 1980 and Doctor Habilitatus in
Technical Sciences in 1990 and became Senior
Researcher in 1984 on Technical Cybernetics and
Information Theory, and Professor in 1994. Areas of
research – computational intelligence, adaptive,
neural, fuzzy, real-time control systems;
identification, diagnostics, fault detection in
technical, economical, medical, ecological systems. |
 |
Dr. Olena Vynokurova, Senior Researcher of
Control Systems Research Laboratory and associate
professor of information technologies security
department in Kharkiv National University of Radio
Electronics (KNURE), Ukraine. She received Master
degree of Computer Intelligent Integrated Systems in
2002 and Ph.D. in Systems and Means of Artificial
Intelligence in KNURE in 2005. Her research
interests are wavelet-, neuro-, fuzzy- hybrid
systems, forecasting, emulation of non-stationary
time sequences, computational intelligence. She is a
member of ITHEA International Scientific Society. |
 |
Dr. Yegorova Elena Senior Researcher of
Computer Science department and associate professor
of Computering Machines department in Kharkiv
National University of Radio Electronics (KNURE),
Ukraine. She received M.Phill in Information
technologies, Data mining and Knowledge discovery in
University of Wales (UK) in 2006 and Ph.D. in
Systems and Means of Artificial Intelligence in
KNURE in 2007. Her research interests are
clusterization, data classification and granulation,
image retrieval, fuzzy systems for medical image
processing. She is a member of ITHEA International
Scientific Society.
|
BibTex
@ARTICLE{P1120827003,
AUTHOR =
{Ye. Bodyanskiy
and O. Vynokurova and E. Yegorova},
TITLE = {RADIAL-BASIS-FUZZY-WAVELET-NEURAL
NETWORK WITH ADAPTIVE ACTIVATION-MEMBERSHIP FUNCTION},
JOURNAL = {ICGST International Journal on Artificial Intelligence and Machine Learning,
AIML},
YEAR = {2008},
VOLUME = {8},
ISSUE ={II},
}
(P1120827003,
971 KB)
|
