Klasifikasi Pasien Penyakit Jantung Koroner Menggunakan Metode Hybrid Support Vector Machine-Genetic Algorithm (SVM-GA) Dan Decision Tree-Genetic Algorithm (DT-GA)

Wibowo, Hammam Prasetyo (2025) Klasifikasi Pasien Penyakit Jantung Koroner Menggunakan Metode Hybrid Support Vector Machine-Genetic Algorithm (SVM-GA) Dan Decision Tree-Genetic Algorithm (DT-GA). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211105-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (4MB) | Request a copy

Abstract

Penyakit Jantung Koroner (PJK) adalah salah satu penyebab utama kematian di dunia, termasuk di Indonesia, dengan prevalensi mencapai 1,5% dan angka kematian tahunan sekitar 245.343 jiwa pada tahun 2024. Deteksi dini sangat penting untuk mengurangi risiko dan dampak dari penyakit ini. Penelitian ini bertujuan untuk mengembangkan model klasifikasi risiko PJK menggunakan pendekatan Support Vector Machine-Genetic Algorithm (SVM-GA) dan Decision Tree-Genetic Algorithm (DT-GA). Pertama, untuk klasifikasi biner tanpa seleksi fitur didapatkan bahwa DT mengungguli SVM dengan accuracy 81,18%, precision 88,46%, recall 75,00%, F1-score 81,16%, dan AUC 89,84%. Hasil klasifikasi tanpa seleksi fitur digunakan untuk membandingkan efektivitas seleksi fitur menggunakan Genetic Algorithm. Setelah itu, GA digunakan untuk memilih subset fitur terbaik dari 13 fitur independen. Metode seleksi fitur berbasis GA membantu mengurangi dimensi fitur yang tidak relevan, meningkatkan efisiensi model, serta mempertahankan fitur yang paling informatif untuk klasifikasi. Hasil penelitian menunjukkan bahwa metode SVM-GA memberikan performa terbaik pada klasifikasi biner dengan accuracy 84,71%, precision 81,13%, recall 93,48%, F1 score 86,67%, dan AUC 93,00%. Sedangkan untuk klasifikasi multikelas, SVM-GA-SMOTE menjadi model terbaik dengan mencapai accuracy 84,84%, macro precision 84,71%, macro recall 84,84%, dan macro F1-score 84,67%, yang menunjukkan kemampuan model dalam mengatasi ketidakseimbangan kelas. Penelitian ini membuktikan bahwa kombinasi SMOTE dan GA efektif meningkatkan performa model, terutama dalam mendeteksi kelas minoritas, dan dapat mendukung pengambilan keputusan klinis dalam deteksi dini PJK.
==================================================================================================================================================
Heart Disease (HD) is one of the leading causes of death worldwide, including in Indonesia, with a prevalence of 1.5% and an annual death toll of approximately 245.343 in 2024. Early detection is crucial to reduce the risk and impact of this disease. This study aims to develop a risk classification model for HD using a Support Vector Machine-Genetic Algorithm (SVM-GA) and Decision Tree-Genetic Algorithm (DT-GA) approach. First, for binary classification without feature selection, it was found that DT outperformed SVM with accuracy 81,18%, precision 88,46%, recall 75,00%, F1-score 81,16%, and AUC 89,84%. The results of the classification without feature selection were used to compare the effectiveness of feature selection using Genetic Algorithm (GA). Afterward, GA was used to select the best feature subset from 13 independent features. The GA-based feature selection method helps reduce the dimensions of irrelevant features, improve model efficiency, and retain the most informative features for classification. The results showed that the SVM-GA method provided the best performance in binary classification with accuracy 84,71%, precision 81,13%, recall 93,48%, F1-score 86,87%, and AUC 93,00%. For multiclass classification, SVM-GA-SMOTE became the best model, achieving accuracy 84,84%, macro precision 84,71%, macro recall 84,84%, and macro F1-score 84,67%, demonstrating the model's ability to address class imbalance. This study proves that the combination of SMOTE and GA effectively improves model performance, especially in detecting minority classes, and can support clinical decision-making for the early detection of HD.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Penyakit Jantung Koroner, Klasifikasi, Decision Tree, Support Vector Machine, Genetic Algorithm, Hybrid Models, Heart Disease, Classification
Subjects:	Q Science Q Science > QA Mathematics Q Science > QA Mathematics > QA402.5 Genetic algorithms. Interior-point methods. Q Science > QA Mathematics > QA76.9.D343 Data mining. Querying (Computer science)
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Hammam Prasetyo Wibowo
Date Deposited:	04 Aug 2025 01:11
Last Modified:	19 Aug 2025 07:43
URI:	http://repository.its.ac.id/id/eprint/126596

Actions (login required)

View Item