cover
Contact Name
ibnu surya
Contact Email
ibnu@pcr.ac.id
Phone
+6285272673321
Journal Mail Official
jurnalkomputerterapan@pcr.ac.id
Editorial Address
Jurnal Komputer Terapan (JKT) Badan Penelitian dan Pengabdian kepada Masyarakat (BP2M) Politeknik Caltex Riau Jl. UmbanSari No. 1 Rumbai - Pekanbaru 28265
Location
Kota pekanbaru,
Riau
INDONESIA
Jurnal Komputer Terapan
Published by Politeknik Caltex Riau
ISSN : 24434159     EISSN : 24605255     DOI : https://doi.org/10.35143/jkt
Core Subject : Science,
Applied Computer Journal Articles from various fields in Informatics, Information Systems and Computer science. Topics included, 1. Informatics 1.1 Software Engineering 1.2 Multimedia 2. Information Systems 2.1 Soft Computing 2.2 Business Analyst 2.3 Data Engineering 3. Computer science 3.1 Operating System 3.2 Computer Network
Articles 223 Documents
OPTIMASI SISTEM INFERENSI FUZZY DENGAN ALGORITMA GENETIKA UNTUK PERSONALISASI REKOMENDASI BEBAN AWAL LATIHAN DEADLIFT Naufal, Atha Redian; Hermawan, Arief
Jurnal Komputer Terapan Vol 11 No 2 (2025): Jurnal Komputer Terapan
Publisher : Politeknik Caltex Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35143/jkt.v11i2.6780

Abstract

The risk of injury in weightlifting, particularly in the deadlift movement, is often caused by subjective initial load determination and exacerbated by the egolifting phenomenon. To mitigate this risk, this research proposes an intelligent hybrid model for personalized load recommendation. This model integrates a Fuzzy Inference System (FIS) with a Genetic Algorithm (GA) to perform end-to-end parameter optimization. The fuzzy system utilizes four inputs (BMI, WHtR, RHR, Experience) to represent the user's condition. The Genetic Algorithm then automatically tunes 18 crucial system parameters, including membership functions and adjustment factors, using 20 real data points from an expert as the ground truth. The research results show that optimization using GA successfully reduced the Mean Absolute Error (MAE) significantly. The validated final model achieved an accuracy of 74.78% and an MAE of 6.37 kg, confirming that the hybrid Fuzzy-Genetic approach is a superior method for tuning quantitative recommendation systems, resulting in more precise and reliable decisions.
PENGEMBANGAN APLIKASI DESKTOP SKINCARE UNTUK REKOMENDASI PRODUK DAN PENJUALAN BERDASARKAN JENIS KULIT Lontaan, Rolly Junius; Saroinsong, Marshanda; Sumual, Monica; Masengi, Mitchelly; Mamuaja, Tiara
Jurnal Komputer Terapan Vol 11 No 2 (2025): Jurnal Komputer Terapan
Publisher : Politeknik Caltex Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35143/jkt.v11i2.6781

Abstract

Skincare products are increasingly used to address various facial skin problems, maintain skin health, and enhance self-confidence. However, the diversity of skin types and the wide range of skincare products available on the market often make it difficult for users to choose the most suitable products. To address this issue, this study developed a desktop-based skincare application equipped with a recommendation system that uses rule-based and weighted scoring methods, implemented using Python and a MySQL database. The recommendation system analyzes users’ skin types and evaluates the compatibility of product ingredients to generate more accurate recommendations. The test results show that the application can provide suitable recommendations for four skin type categories (dry, oily, sensitive, and combination) with a recommendation suitability level of 84% based on user evaluation. In addition, the application offers a product sales feature that facilitates the process from product selection to purchase. Therefore, this application serves as an effective digital solution to help users determine the right skincare products according to their specific skin needs.
ANALISIS FAKTOR DOMINAN KEBERHASILAN AKADEMIK MAHASISWA PENERIMA BEASISWA MENGGUNAKAN PRINCIPAL COMPONENT ANALYSIS Mufidah, Nur; Syarif Sihabudin Sahid, Dadang; Perdana Arifin, Satria; Ari Sandi, Wahyu
Jurnal Komputer Terapan Vol 11 No 2 (2025): Jurnal Komputer Terapan
Publisher : Politeknik Caltex Riau

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35143/jkt.v11i2.6819

Abstract

The academic success of scholarship recipients is not only determined by their academic abilities but also by various non-academic factors that are often difficult to measure objectively. The large number of interrelated indicators presents a challenge in identifying the most dominant factors, especially when these variables exhibit high correlations and lead to data redundancy. This study aims to identify the dominant non-academic factors influencing the academic success of scholarship recipients using Principal Component Analysis (PCA). A total of 262 students completed a questionnaire consisting of 54 non-academic items that had previously undergone validity and reliability testing. PCA was employed to reduce data dimensionality and produced 38 principal components with a cumulative explained variance of 95.08%, indicating effective dimensionality reduction without significant loss of information. The loading matrix analysis revealed that psychological conditions, learning methods, major suitability, learning motivation, and financial conditions were the most dominant contributors to the principal components. These findings provide a more structured understanding of the non-academic factors that should be considered in the development and monitoring of scholarship programs.