Claim Missing Document
Check
Articles

Found 14 Documents
Search

Perancangan Aplikasi Augmented Reality 3D Sebagai Media Pembelajaran Rumah Adat Indonesia Dengan Algoritma FAST Corner di SDI Al-Munir Bekasi Widyanto, Abdillah Prayoga; Handayani, Dwipa; Mayadi
Journal of Students‘ Research in Computer Science Vol. 6 No. 1 (2025): Mei 2025
Publisher : Program Studi Informatika Fakultas Ilmu Komputer Universitas Bhayangkara Jakarta Raya

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31599/bkc3ne60

Abstract

Elementary school students still tend to be less interested in learning local culture, one of which is the variety of traditional houses in Indonesia. Students are more interested in global culture that is trending. The main reason is because the learning process that applies is still conventional and monotonous. The teaching methods that are carried out are only delivered verbally and through written media, resulting in low student involvement in the learning process. Many areas that have cultural heritage are at risk of losing valuable cultural assets due to limited interest and preservation efforts. Digital transformation allows for the creation of new innovations in the scope of education. Augmented reality technology can be used as a bridge to connect students with local cultural diversity. AR technology designed on mobile devices is an ideal medium to motivate the younger generation to maintain the sustainability of local culture in a modern way and in accordance with a digital lifestyle. The purpose of this study is to explore the use of 3D Augmented Reality technology collaborated with the FAST corner algorithm in the learning process as an effort to increase student interest in understanding the forms of traditional houses in Indonesia. The development of a waterfall system is a method in developing Android-based applications by utilizing Unity 3D programming. This research produces an Android-based application to recognize various forms of traditional houses in Indonesia which will be applied to 3rd grade elementary school students so that they can experience a more interactive learning experience.
Pelatihan dan Pendampingan Penggunaan Sistem Prediksi Kelulusan Mahasiswa dengan Fuzzy C-Means Mayadi
Jurnal Ilmiah Pengabdian dan Inovasi Vol. 3 No. 4 (2025): Jurnal Ilmiah Pengabdian dan Inovasi (Juni)
Publisher : Insan Kreasi Media

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.57248/jilpi.v3i4.648

Abstract

The problem of low on-time graduation rates in private universities is a major challenge in academic management and institutional accreditation processes. This service activity aims to provide training and assistance to academic managers at Bumigora University in the use of a student graduation prediction system based on the Fuzzy C-Means method. This activity includes training in the use of data mining, understanding algorithms, and implementation using the Python programming language. This service activity includes four stages: (1) identification of partners and needs, namely academic management personnel of Bumigora University who need a graduation prediction system and data analysis skills; (2) training and workshops on data mining, Fuzzy C-Means, and CRISP-DM with hands-on practice using Python; (3) system implementation through simulation of student graduation prediction; and (4) evaluation through pre-test, post-test, and reflective discussion. All stages are integrated to improve the capacity of academic management. The results showed an increase in participants' understanding and ability to operate the system to predict student graduation. This system is expected to help the academic monitoring process more effectively and data-based.
Pelatihan Pembuatan Sistem Informasi Penjualan Produk Tani Untuk Kelompok Tani Muda di Desa Labulia Rosanensi, Melati; Sakti, Lanang; Santoso, Heroe; Madani, Miftahul; Suriyati; Mayadi
Jurnal Ilmiah Pengabdian dan Inovasi Vol. 3 No. 2 (2024): Jurnal Ilmiah Pengabdian dan Inovasi (Desember)
Publisher : Insan Kreasi Media

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.57248/jilpi.v3i2.524

Abstract

As time goes by, information systems are still very much needed, both for individuals and groups. Information systems are still very reliable, especially for sales, such as selling agricultural products to Indonesian young farmers' associations. The Indonesian Young Farmers Association is a community that sells and produces various agricultural products such as fertilizers, pesticides, medicines, seeds, fruits, vegetables, spices, and food crops. The Indonesian Young Farmers Association is one of the farmer groups located in rural areas. So far the Indonesian Young Farmers Association still uses a conventional system, namely customers who come to their place of business, so buyers have to go to the place where they buy agricultural products and customers need energy to choose the goods or agricultural products they want to buy, and to face intense competition they must have a marketing strategy. Different from its competitors, this requires an information system that can provide information about products to customers quickly and accurately. This research uses the waterfall method which starts from analysis, design, implementation, testing, and maintenance. Results From the trial, a tabulation of 100% was obtained which stated that they strongly agreed that the system that had been created was suitable and could be implemented in the Farmer Product Sales Information System at the Indonesian Young Farmers Association for village progress.
Penerapan Ensemble Learning dengan Hard Voting untuk Klasifikasi Customer Churn Astawa, Andhika rama putra; Martono, Galih Hendro; Mayadi
CORISINDO 2025 Vol. 1 (2025): Prosiding Seminar Nasional CORISINDO 2025
Publisher : CORISINDO 2025

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30812/corisindo.v1.5340

Abstract

Customer churn menjadi salah satu tantangan terbesar bagi perusahaan telekomunikasi karena berdampak langsung pada pendapatan dan keberlanjutan bisnis. Penelitian ini bertujuan untuk meningkatkan akurasi prediksi churn dengan mengembangkan model ensemble learning berbasis Hard Voting Classifier yang menggabungkan tiga algoritma berbeda, yaitu Naïve Bayes, Random Forest, dan Nearest Centroid. Dataset pelanggan yang digunakan mencakup informasi demografis, perilaku penggunaan layanan, dan status churn, yang kemudian diproses melalui tahapan pembersihan data, seleksi fitur, normalisasi, serta teknik resampling SMOTE-Tomek untuk menyeimbangkan distribusi kelas. Pemilihan fitur dilakukan dengan metode Information Gain dan analisis korelasi, sehingga hanya atribut yang relevan digunakan dalam pemodelan. Hasil pengujian menunjukkan bahwa Hard Voting Classifier mampu mencapai akurasi sebesar 90% dengan nilai recall untuk kelas churn sebesar 81%, lebih tinggi dibandingkan Random Forest (78%), meskipun akurasi Random Forest lebih tinggi (95%). Nilai precision untuk kelas non-churn juga meningkat hingga 97%, menandakan model ini efektif mengurangi kesalahan dalam memprediksi pelanggan tetap. Temuan ini membuktikan bahwa pendekatan ensemble learning dengan base learner heterogen dapat memadukan keunggulan masing-masing algoritma untuk meningkatkan deteksi churn. Meski demikian, performa Hard Voting masih bergantung pada kualitas masing-masing classifier, sehingga optimasi hyperparameter dan eksplorasi kombinasi model lain direkomendasikan untuk penelitian selanjutnya. Hasil penelitian ini diharapkan dapat membantu perusahaan merumuskan strategi retensi pelanggan yang lebih tepat sasaran dan berkelanjutan.