Claim Missing Document
Check
Articles

Found 2 Documents
Search

Implementasi Metode Case-Based Reasoning (CBR) dalam Sistem Pakar untuk Mendapatkan Diagnosis Anxiety Disorders Gunung, Tar Muhammad Raja; Lubis, Siti Sahara; Siregar, Manutur; Simanjuntak, Peter Jaya Negara; Jinan, Abwabul
Jurnal Teknologi Terpadu Vol 10 No 2 (2024): Desember, 2024
Publisher : LPPM STT Terpadu Nurul Fikri

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54914/jtt.v10i2.1480

Abstract

This research aims to develop an expert system based on the case-based reasoning method for diagnosing anxiety disorders. Anxiety Disorder is a mental health disorder that is often experienced by the public but is often not detected correctly. The case-based reasoning method was chosen because of its ability to utilise previous cases to solve new problems that have similarities. Case-based reasoning uses four main stages: retrieval, reuse, revise, and retain. The case-based reasoning method is implemented using case data obtained from psychology clinics and interviews with mental health experts. Testing the case-based reasoning method shows a high level of accuracy in diagnosing various types of Anxiety Disorders, such as Generalised Anxiety Disorder, Panic Disorder, and Specific Phobias. The results of this study show that the case-based reasoning method can be an effective tool in helping mental health professionals diagnose Anxiety Disorders more quickly and accurately. After searching using the symptoms obtained, the percentage of each type of disease is the percentage of Generalised Anxiety Disorder 35.7%, the percentage of Panic Disorder 30.7%, and the percentage of Specific Phobias 65%.
ANALISIS PERBANDINGAN RANDOM FOREST DAN KNN PADA KLASIFIKASI PENERIMA MANFAAT PROGRAM MAKAN BERGIZI GRATIS Zarkasyi, Muhammad Imam; Simanjuntak, Peter Jaya Negara; Nababan, Junerdi
JOURNAL OF SCIENCE AND SOCIAL RESEARCH Vol 9, No 1 (2026): February 2026
Publisher : Smart Education

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54314/jssr.v9i1.5753

Abstract

Abstract: The Free Nutritional Meal Program (PMBG) is a government initiative to improve students' nutrition and school attendance. This study evaluates and compares the performance of Random Forest (RF) and K-Nearest Neighbors (KNN) algorithms in classifying students most eligible for PMBG based on socio-economic criteria. The dataset comprises 205 public elementary school students in Medan City, collected via questionnaires. Features include parental income, number of dependents, housing status, asset ownership, and participation in other social aid programs. The data was clustered into three priority groups using K-Means. To address class imbalance, the Synthetic Minority Over-sampling Technique (SMOTE) was applied. Modeling used three data split scenarios (70:30, 80:20, 90:10) and was evaluated with accuracy, precision, recall, F1-score, and cross-validation. Results show that RF consistently outperformed KNN across all scenarios. After SMOTE, both models improved, with Balanced-RF achieving the highest accuracy and F1-score (94%) in the 70:30 split. The combination of RF and SMOTE proves effective for building an objective and accurate priority classification system for PMBG beneficiaries. Keyword: Free Nutritious Meal Program; Random Forest; K-Nearest Neighbors; SMOTE. Abstrak: Program Makan Bergizi Gratis (PMBG) merupakan inisiatif pemerintah yang bertujuan meningkatkan asupan gizi dan mendorong kehadiran siswa di sekolah. Penelitian ini bertujuan untuk mengevaluasi dan membandingkan kinerja dua algoritma machine learning, yaitu Random Forest dan K-Nearest Neighbors, dalam mengklasifikasikan siswa yang paling berhak menerima manfaat PMBG berdasarkan kriteria sosial-ekonomi. Dataset yang digunakan terdiri dari 205 siswa Sekolah Dasar Negeri di Kota Medan yang dikumpulkan melalui kuesioner. Fitur yang digunakan meliputi pendapatan orang tua, jumlah tanggungan, status tempat tinggal, kepemilikan aset, dan partisipasi dalam program bantuan sosial lainnya. Dataset yang telah dikumpulkan kemudian dikelompokkan menggunakan algoritma K-Means menjadi tiga klaster prioritas. Untuk mengatasi ketidakseimbangan distribusi data, digunakan metode Synthetic Minority Over-sampling Technique (SMOTE). Pemodelan dilakukan dalam tiga skenario pembagian data (70:30, 80:20, 90:10) dan dievaluasi menggunakan metrik akurasi, presisi, recall, f1-score, dan cross-validation. Hasil penelitian menunjukkan bahwa algoritma Random Forest secara konsisten memberikan kinerja yang lebih unggul dibandingkan KNN pada semua skenario. Setelah penerapan SMOTE, kedua algoritma mengalami peningkatan performa, dengan Random Forest-Balanced mencatat akurasi dan f1-score tertinggi sebesar 94% pada skenario 70:30. Temuan ini menunjukkan bahwa kombinasi Random Forest dan SMOTE merupakan pendekatan yang efektif dan efisien untuk membangun sistem klasifikasi prioritas penerima manfaat PMBG yang objektif dan akurat. Kata kunci: Program Makan Bergizi Gratis; Random Forest; K-Nearest Neighbors; SMOTE