Open AccessThe Klasifikasi Jenis Ikan Berbasis Jaringan Saraf Tiruan Menggunakan Algoritma Principal Component Analysis (PCA)
Author(s) -
Arif Lumute Unihehu,
Imam Suharjo
Publication year - 2021
Publication title -
jurnal ilmiah ilmu komputer fakultas ilmu komputer universitas al asyariah mandar/jurnal ilmiah ilmu komputer
Language(s) - English
Resource type - Journals
eISSN - 2503-3832
pISSN - 2442-451X
DOI - 10.35329/jiik.v7i2.200
Subject(s) - principal component analysis , fish <actinopterygii> , pattern recognition (psychology) , artificial intelligence , biology , computer science , fishery
Fish are cold-blooded animals that are widely used by humans. Fish are a diverse group of poikilothermic vertebrates with more than 27,000 species worldwide. A large number of fish species becomes a problem in distinguishing the types of fish. The purpose of this study was to create a fish type classification system based on the texture of artificial neural network-based fish imagery using K-Nearest Neighbors and Principal Component Analysis (PCA) algorithms. The data was taken through direct exploration and retrieved directly by researchers. The data only uses 3 types of fish as the object of further research conducted training and testing test data in the first, second, and third classes only one can not be recognized by the system, while the other data can be recognized by the percentage of success of 93% (Ninety-three percent).