Napredna pretraga

Pregled bibliografske jedinice broj: 763879

Automatic Design of Radial Basis Function Networks Through Enhanced Differential Evolution


Bajer, Dražen; Zorić, Bruno; Martinović, Goran
Automatic Design of Radial Basis Function Networks Through Enhanced Differential Evolution // Proceedings of the 10th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2015, Lecture Notes in Computer Science, Vol. 9121 / Onieva, Enrique ; Santos, Igor ; Osaba, Eneko ; Quintián, Héctor ; Corchado, Emilio (ur.).
Bilbao, Španjolska: Springer International Publishing Switzerland, 2015. str. 244-256 (predavanje, međunarodna recenzija, cjeloviti rad (in extenso), znanstveni)


Naslov
Automatic Design of Radial Basis Function Networks Through Enhanced Differential Evolution

Autori
Bajer, Dražen ; Zorić, Bruno ; Martinović, Goran

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

Izvornik
Proceedings of the 10th International Conference on Hybrid Artificial Intelligence Systems, HAIS 2015, Lecture Notes in Computer Science, Vol. 9121 / Onieva, Enrique ; Santos, Igor ; Osaba, Eneko ; Quintián, Héctor ; Corchado, Emilio - : Springer International Publishing Switzerland, 2015, 244-256

ISBN
978-3-319-19643-5

Skup
International Conference on Hybrid Artificial Intelligence Systems

Mjesto i datum
Bilbao, Španjolska, 22-24.06.2015

Vrsta sudjelovanja
Predavanje

Vrsta recenzije
Međunarodna recenzija

Ključne riječi
Differential evolution; Initial population; k-means; Neural network; Radial basis function

Sažetak
During the creation of a classification model, it is vital to keep track of numerous parameters and to produce a model based on the limited knowledge inferred often from very confined data. Methods which aid the construction or completely build the classification model automatically, present a fairly common research interest. This paper proposes an approach that employs differential evolution enhanced through the incorporation of additional knowledge concerning the problem in order to design a radial basis neural network. The knowledge is inferred from the unsupervised learning procedure which aims to ensure an initial population of good solutions. Also, the search space is dynamically adjusted i.e. narrowed during runtime in terms of the decision variables count. The results obtained on several datasets suggest that the proposed approach is able to find well performing networks while keeping the structure simple. Furthermore, a comparison with a differential evolution algorithm without the proposed enhancements and a particle swarm optimization algorithm was carried out illustrating the benefits of the proposed approach.

Izvorni jezik
Engleski

Znanstvena područja
Računarstvo



POVEZANOST RADA


Projekt / tema
165-0362980-2002 - Postupci raspoređivanja u samoodrživim raspodijeljenim računalnim sustavima (Goran Martinović, )

Ustanove
Fakultet elektrotehnike, računarstva i informacijskih tehnologija Osijek

Časopis indeksira:


  • Scopus