Naslov
Adaptive Polynomial Neural Networks for Times Series Forecasting

Autori
Liatsis, Panos ; Foka, Amalia ; Goulermas, John Yannis ; Mandić, Lidija

Vrsta, podvrsta i kategorija rada
Radovi u zbornicima skupova, cjeloviti rad (in extenso), znanstveni

Izvornik
Proceedings ELMAR-2007 / Grgić, M ; Grgić, S. - Zagreb : Hrvatsko društvo Elektronika u pomorstvu (ELMAR), 2007, 35-39

ISBN
978-953-7044-05-3

Skup
49th International Symposium ELMAR-2007

Mjesto i datum
Zadar, Hrvatska, 12.09.2007. - 14.09.2007

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Genetic Algorithms; polynomial neural networks; time series; forecasting
(genetic algorithms; polynomial neural networks; time series; forecasting)

Sažetak
Time series prediction involves the determination of an appropriate model, which can encapsulate the dynamics of the system, described by the sample data. Previous work has demonstrated the potential of neural networks in predicting the behaviour of complex, non-linear systems. In particular, the class of polynomial neural networks has been shown to possess universal approximation properties, while ensuring robustness to noise and missing data, good generalisation and rapid learning. In this work, a polynomial neural network is proposed, whose structure and weight values are determined with the use of evolutionary computing. The resulting networks allow an insight into the relationships underlying the input data, hence allowing a qualitative analysis of the models´ performance. The approach is tested on a variety of non-linear time series data.

Izvorni jezik
Engleski

Znanstvena područja
Elektrotehnika

POVEZANOST RADA