Claim Missing Document
Check
Articles

Found 2 Documents
Search

APLIKASI REGISTRASI PENERIMAAN MAGANG ONLINE PADA BANK JAMBI Lestari, Dewi; Bintana, Rizqa Raaiqa; Budiman, Naufal
ScientiCO : Computer Science and Informatics Journal Vol 3, No 2 (2020): Scientico : November
Publisher : Fakultas Teknik, Universitas Tadulako

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Internship is an activity carried out by students to plunge directly into the world of work in order to increase experience and improve student abilities as well as a forum for implementing knowledge gained from higher education. PT. Regional Development Bank (BPD) of Jambi City or better known as Bank Jambi is an agency that acts as a development bank, commercial bank, regional cash holder and is one of the sources of regional income. As one of the agencies that provides opportunities for students to be able to carry out internship activities, the registration process and internships acceptance at Bank Jambi are still done manually. In today's increasingly developing era, the registration process can actually be carried out without having to come directly to Bank Jambi to deliver a letter and then wait for confirmation whether students are accepted for an internship at Bank Jambi or not. Likewise, the internship student registration file storage system is still done manually, so this is very inefficient, therefore with this online internship registration aplication it is hoped that it can improve the efficiency of the internship acceptance process at Bank Jambi, as a consideration in taking a decision and as part of information technology-based innovation. Making this aplication using the waterfall method, for the analysis and design stages using an object-oriented methodology using the UML diagram tool. For the analysis phase using activity diagrams and use case diagrams, for the design stage using class diagrams.
Optimasi Convolutional Neural Network Menggunakan Differential Evolution dalam Identifikasi Kematangan Buah Kelapa Sawit Budiman, Naufal; Adi, Kusworo; Wibowo, Adi
Jurnal Teknologi Informasi dan Ilmu Komputer Vol 13 No 1: Februari 2026
Publisher : Fakultas Ilmu Komputer, Universitas Brawijaya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.25126/jtiik.2026131

Abstract

Deteksi kematangan buah kelapa sawit merupakan salah satu langkah penting dalam meningkatkan efisiensi dan produktivitas di sektor pertanian di Indonesia. Dalam era perkembangan teknologi, penerapan metode berbasis kecerdasan buatan, seperti Convolutional Neural Network (CNN), sering digunakan dalam pengenalan dan klasifikasi citra. Dalam proses pengembangan model Deep Learning, optimasi untuk meningkatkan akurasi dan efisiensi komputasi menjadi langkah penting. Proses ini melibatkan penyesuaian arsitektur dan hyperparameter untuk memastikan model dapat mempelajari fitur yang relevan secara efektif. Melalui optimasi, model dapat disesuaikan untuk menangani karakteristik dataset tertentu, meningkatkan kemampuan generalisasi, dan memaksimalkan kinerja pada tugas klasifikasi. Penelitian ini mengoptimasi model CNN melalui penyesuaian arsitektur dan hyperparameter menggunakan salah satu algoritma evolusi, yaitu Differential Evolution (DE), dalam mengidentifikasi tingkat kematangan buah kelapa sawit. Dataset yang digunakan dalam penelitian ini berupa gambar buah kelapa sawit dengan tiga tingkat kematangan, yaitu: matang, mentah, dan busuk. Sebagai pembanding, penelitian ini juga menggunakan metode CNN tanpa DE. Hasil dari penelitian menunjukkan bahwa metode CNN yang dioptimasi menggunakan DE dalam mengidentifikasi tingkat kematangan buah kelapa sawit memberikan akurasi sebesar 0,98 serta nilai presisi, sensitivitas, dan F1-score di atas 0,97 untuk semua kelas.   Abstract The detection of oil palm fruit ripeness is a crucial step in improving efficiency and productivity in Indonesia's agricultural sector. In the era of technological advancement, artificial intelligence-based methods, such as Convolutional Neural Networks (CNN), are frequently applied in image recognition and classification. In developing Deep Learning models, optimization plays a vital role in enhancing accuracy and computational efficiency. This process involves adjusting the architecture and hyperparameters to ensure the model can effectively learn relevant features. Through optimization, the model can be tailored to handle specific dataset characteristics, improve generalization, and maximize performance in classification tasks. This study optimizes a CNN model by adjusting its architecture and hyperparameters using an evolutionary algorithm known as Differential Evolution (DE) to identify the ripeness levels of oil palm fruits. The dataset used in this study consists of images of oil palm fruits categorized into three ripeness levels: ripe, unripe, and rotten. For comparison, a baseline CNN model without DE optimization was also employed. The results show that the CNN model optimized using DE achieved an accuracy of 0,98 with precision, sensitivity, and F1-score values exceeding 0.97 for all classes.