Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal PERFORMA : Media Ilmiah Teknik Industri
Yusuf Priyandari
Program Studi Teknik Industri Fakultas Teknik, Universitas Sebelas Maret
Engineering Industrial & Manufacturing Engineering

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
Implementasi Deep Learning Menggunakan Metode Convolutional Neural Network untuk Mendeteksi Kehalaln pada Kosmetik Hafsah Qonita; Pringgo Widyo Laksono; Yusuf Priyandari
Performa: Media Ilmiah Teknik Industri Vol 22, No 1 (2023): Performa: Media Ilmiah Teknik Industri
Publisher : Industrial Engineering, Faculty of Engineering, Universitas Sebelas Maret

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20961/performa.22.1.76650

Abstract

The growth rate of the cosmetics industry shows good development, reaching 9.39% in 2020. With so many cosmetic products in the market, consumers must be more careful in choosing the cosmetic products. In addition to the safety factor, the halalness of cosmetics also needs to be considered, especially for Muslim consumers. This research aims to create a halal detection model in cosmetics by implementing one of the deep learning methods, namely convolutional neural network (CNN). .Previous research has successfully created a halal detection model on Korean cosmetics using CNN with an accuracy rate of 95.56%. This research intends to develop previous research by adding classes and the number of datasets. CNN will be used to create a halal detection model in cosmetics by learning the input features in the form of the image of cosmetic ingredient to determine its halalness. Classification is done based on two classes, which are Halal and Shubhat. The results show that the CNN model gets an accuracy value of 98.66% with a loss of 0.0615 in classifying the halalness of cosmetics. Model testing using the testing dataset gets an accuracy value of 98.67%. The F1-score value in each class is 98.66% for the halal class and 98.67 for the shubhat class. The CNN model that has been created is appropriate because it shows high accuracy and low loss on training, validation, and testing data without experiencing overfitting or underfitting,
Co-Authors Hafsah Qonita Pringgo Widyo Laksono