Claim Missing Document
Check
Articles

Found 23 Documents
Search

Analysis Of Decision Support Systems Edas Method In New Student Admission Selection Siregar, Yunita Sari; Zakir, Ahmad; Syahputri, Nenna Irsa; Harahap, Herlina; Handoko, Divi
Journal of Computer Networks, Architecture and High Performance Computing Vol. 5 No. 1 (2023): Article Research Volume 5 Issue 1, January 2023
Publisher : Information Technology and Science (ITScience)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47709/cnahpc.v5i1.2057

Abstract

University of Harapan Medan is one of the private tertiary institutions in North Sumatra which has an informatics engineering study program. The informatics engineering study program is a study program that has many enthusiasts. Every year this study program graduates more than 200 students. To produce graduates who have potential, reliability and competence in the field of technology and information, it is necessary to make a selection at the beginning, namely at the time of admission of new students. There are 5 criteria used in the selection process, including the average report card score, basic ability test, computer ability test, psychological test, and interview. Each criterion has 5 weights of values, namely very high, high, medium, low and very low.  The selection process for admission of new informatics engineering students with a decision support system for the EDAS (Evaluation Based On Distance From Average Solution) method.  Where the stages in this method are by normalizing the decision matrix and looking for the average from alternatives, then from these results calculate the average positive distance (PDA) and negative distance (NDA) as well as the assessment of the weighted attribute weights of SPi and SNi, after that the normalization of positive and negative distance weights is carried out for determining the ranking score. From the results of the analysis carried out using the EDAS method, with a sample of 10 prospective students it was concluded that the 6th order student candidate had the highest score with a score of 0.519 and the lowest score in the 7th order student with a score of 0.14. Therefore, the level of accuracy of the EDAS method in selecting new student admissions is around 20%. Of course, this accuracy value will change with large data samples.
APPLICATION OF PROTOTYPE METHOD IN PERIODIC SERVICE INFORMATION SYSTEM APPLICATION (Case Study: MELET TUNING WORKSHOP MEDAN) Ramadini, Dwi Wulan Indah; Zakir, Ahmad; Sembiring, Boni Oktaviana
Bahasa Indonesia Vol 16 No 06 (2025): Instal : Jurnal Komputer
Publisher : Cattleya Darmaya Fortuna

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54209/jurnalinstall.v16i06.323

Abstract

This study aims to build a website-based periodic service information system application, By applying the prototype method application development method through concepts, design, and evaluation provides solutions to problems. It was found that the use of periodic service applications in the Melet Tuning Medan workshop, which has been built is very useful because of the results of the application of methods and system design. can create satisfaction with the workshop's services to customers. get service and the workshop can manage service data well and efficiently. This study produces a prototype design and applies methods in building periodic service applications on the website system.
Deteksi Malware Berbasiskan Analisis Statis Menggunakan Algoritma Convolutional Neural Network (CNN) Wiyono, Tri; Hadinata, Edrian; Zakir, Ahmad; Elhanafi, Andi Marwan
Jurnal Sistem Komputer dan Informatika (JSON) Vol. 7 No. 1 (2025): September 2025
Publisher : Universitas Budi Darma

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.30865/json.v7i1.8946

Abstract

Sistem informasi berperan penting dalam menjaga kelancaran operasional organisasi sekaligus melindungi data dari ancaman siber, termasuk malware. Pertumbuhan akses internet di Indonesia yang mencapai lebih dari 215 juta pengguna pada 2023 dan penetrasi smartphone Android sebesar 132,7 juta pada 2025 meningkatkan potensi risiko keamanan. Rendahnya kesadaran pengguna Android terhadap keamanan informasi (rata-rata 37%) menjadi faktor dominan munculnya insiden spam, phishing, dan malware. Dalam konteks ini, analisis statis menjadi salah satu pendekatan efektif untuk mendeteksi malware melalui izin aplikasi (permissions) yang tertera pada file AndroidManifest.xml. Namun, kelemahannya terletak pada keterbatasan mendeteksi kode berteknik obfuscation. Penelitian ini mengembangkan model deteksi malware berbasis Convolutional Neural Network (CNN) dengan memanfaatkan fitur statis APK. Dataset yang digunakan mencakup aplikasi berbahaya seperti Undangan Pernikahan.apk serta data publik dari Kaggle. Proses meliputi ekstraksi izin aplikasi, normalisasi, pembentukan vektor 256 fitur, serta klasifikasi biner malware dan benignware. Hasil uji menunjukkan akurasi 92% dengan precision tinggi (0,93) namun recall pada malware relatif rendah (0,77), mengindikasikan masih adanya false negative signifikan. Temuan ini menegaskan bahwa CNN efektif untuk deteksi berbasis izin, tetapi peningkatan recall diperlukan agar sistem lebih andal. Pengembangan pendekatan hibrida dengan menggabungkan analisis statis dan dinamis disarankan untuk memperkuat deteksi malware Android yang semakin kompleks. Dengan demikian, penelitian ini diharapkan memberikan kontribusi signifikan terhadap pengembangan sistem keamanan siber yang adaptif, andal, serta relevan bagi tantangan digital masa depan.