APLIKASI JARINGAN SYARAF TIRUAN UNTUK PENDUGAAN MUTU MANGGA SEGAR SECARA NON-DESTRUKTIF
ABSTRACT: The objective of
this study was to estimate grade of mango with artificial neural network. There are three grade of mango: grade I,
grade II and grade III. Each grade consists of 40 samples. The Artificial
Neural Network consists of three model; first model: 3 input, second model: 5
input and third model: 7 input. Each of
models consists of three outputs with 1000, 2500, and 5000 iteration, and 3, 5,
and 7 hidden layers. The result of this study show that the accuracy of
artificial neural network in grade of mango: (1) the first model 60% -
100%; (2) the second model 90% and (3)
the third model 90 %. The conclusion of this study is the artificial neural
network can predict the grade of mango and the best model is the third model.
Keywords: Mango, grade,
artificial neural network
Penulis: Sandra Sandra
Kode Jurnal: jppertaniandd050025
