PAMUNGKAS, Ardian and Isnanto, R. Rizal and Mutiara K.N., Dinar (2025) IMPLEMENTASI K-NEAREST NEIGHBOR DALAM CASE BASED REASONING PADA SISTEM DIAGNOSIS KESEHATAN MENTAL. Masters thesis, UNIVERSITAS DIPONEGORO.

	Text COVER1.pdf Download (298kB)
	Text COVER LENGKAP.pdf Restricted to Repository staff only Download (3MB)
	Text BAB I.pdf Download (276kB)
	Text BAB II.pdf Download (663kB)
	Text BAB III.pdf Restricted to Repository staff only Download (816kB)
	Text BAB IV.pdf Restricted to Repository staff only Download (1MB)
	Text BAB V.pdf Restricted to Repository staff only Download (269kB)
	Text DAFTAR PUSTAKA.pdf Download (289kB)
	Text LAMPIRAN.pdf Restricted to Repository staff only Download (1MB)

Abstract

Penelitian ini berfokus pada pengembangan sistem diagnosis kesehatan mental menggunakan metode CBR dan algoritma KNN. Tujuan dari penelitian ini adalah untuk mengevaluasi performa model diagnosis kesehatan mental berbasis metode KNN dalam pendekatan CBR, melalui analisis akurasi, ketepatan klasifikasi, dan ketanggapan model terhadap variasi data kasus, serta merancang dan mengembangkan model diagnosis yang adaptif dengan mengintegrasikan metode KNN ke dalam kerangka CBR guna mengkaji kontribusinya terhadap peningkatan efektivitas dan akurasi sistem pendukung keputusan di bidang kesehatan mental. Manfaat yang diharapkan dari penelitian ini meliputi peningkatan efektivitas layanan diagnosis kesehatan mental dan kontribusi terhadap literatur akademik di bidang ini. Metode penelitian melibatkan prapemrosesan dengan Text Mining, perhitungan TF-IDF dan untuk mengatasi ketidakseimbangan kelas dalam dataset digunakan teknik oversampling SMOTE. Cosine Similarity diterapkan untuk mengukur kesamaan antar data, dan model KNN digunakan untuk klasifikasi dengan pemilihan parameter optimal melalui penentuan berdasar nilai ganjil. Evaluasi model dilakukan menggunakan Confusion Matrix, menghasilkan akurasi 91,67% pada data uji, serta metrik evaluasi lainnya seperti precision, recall, dan F1-score. Hasil penelitian menunjukkan bahwa penerapan KNN dalam CBR dapat memberikan diagnosis yang baik dalam mengenali kasus kesehatan mental dari data teks.
Kata Kunci: Case Based Reasoning, Diagnosis Kesehatan Mental, K-Nearest Neighbor

This study focuses on the development of a mental health diagnosis system using the CBR method and the KNN algorithm. The objective of this research is to evaluate the performance of a mental health diagnosis model based on the KNN method within the CBR framework, through the analysis of accuracy, classification precision, and the model’s responsiveness to variations in case data, as well as to design and develop an adaptive diagnosis model by integrating KNN into the CBR framework to examine its contribution to improving the effectiveness and accuracy of decision support systems in the field of mental health. The expected benefits of this research include enhancing the effectiveness of mental health diagnostic services and contributing to academic literature in this domain. The research methodology involves preprocessing with Text Mining, feature extraction using TF-IDF, and handling class imbalance in the dataset using the SMOTE oversampling technique. Cosine Similarity is applied to measure similarity between data points, and the KNN model is used for classification with optimal parameter selection based on odd-numbered values of k. Model evaluation is conducted using a Confusion Matrix, resulting in an accuracy of 91.67% on the test data, along with other evaluation metrics such as precision, recall, and F1-score. The findings indicate that applying KNN within the CBR framework can provide reliable diagnosis in identifying mental health conditions from textual data.
Keywords: Case Based Reasoning, Mental Health Diagnosis, K-Nearest Neighbor

Actions (login required)

View Item