TELKOMNIKA Telecommunication, Computing, Electronics and Control
Metaheuristic optimization in neural network model for seasonal data

TELKOMNIKA Telecommunication, Computing, Electronics and Control
Metaheuristic optimization in neural network model for seasonal data

Metaheuristic
Neural network
Optimization
Rainfall
Time series

The use of metaheuristic optimization techniques in obtaining the optimal
weights of neural network model for the time series was the main part of this
research. The three optimization methods used as experiments were genetic
algorithm (GA), particle swarm optimization (PSO), and modified bee colony
(MBC). Feed forward neural network (FFNN) was the neural network (NN)
architecture chosen in this research. The limitations and weaknesses of
gradient-based methods for learning algorithm inspired some researchers to
use other techniques. A reasonable choice is non-gradient based method.
Neural network is inspired by the characteristics of creatures. Therefore, the
optimization techniques which are also resemble the patterns of life in nature
will be appropriate. In this study, various scenarios on the three metaheuristic
optimization methods were applied to get the best one. The proposed
procedure was applied to the rainfall data. The experimental study showed that
GA and PSO were recommended as optimization methods at FFNN model for
the rainfall data.

Budi Warsito, Rukun Santoso, Hasbi Yasin

http://journal.uad.ac.id/index.php/TELKOMNIKA

Sep 15, 2021

peri irawan

pdf

english

text