Identification of sea urchins in melonguane coastal area using Multilayer Perceptron Neural Network
| Autor: | Andar Alwein Pinilas, Luther Alexander Latumakulita, Djoni Hatidja |
|---|---|
| Jazyk: | English<br />Indonesian |
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Ilkom Jurnal Ilmiah, Vol 14, Iss 2, Pp 169-177 (2022) |
| Druh dokumentu: | article |
| ISSN: | 2087-1716 2548-7779 |
| DOI: | 10.33096/ilkom.v14i2.1208.169-177 |
| Popis: | Sea urchins (Echinoidea) are marine biota that is found in Indonesian waters and there are 950 types of sea urchins scattered throughout the world. This study aims to classify types of sea urchins based on the characteristics contained in sea urchin images using the Multilayer Perceptron Neural Network (MLP-NN) method with 3 classification classes. 120 sea urchin image data were taken from the Melonguane beach area, Talaud Islands Regency, North Sulawesi Province. In the MLP-NN stage, training, validation, and testing processes are carried out by applying 8-fold cross-validation, and the system performance shows the lowest accuracy of 93.33%, the highest 100%, and an average of 98.33%. The experimental results indicate that MLP-NN can classify sea urchins with good performance. |
| Databáze: | Directory of Open Access Journals |
| Externí odkaz: |
|