Prediksi Penyakit Tuberculosis Menggunakan Algoritma Naïve Bayes dan Random Forest

Abstrak

Abstrak
Tuberkulosis (TBC) merupakan salah satu penyakit menular yang masih menjadi permasalahan kesehatan global, termasuk di Indonesia. Tingginya angka kasus TBC mendorong perlunya sistem prediksi yang mampu membantu proses deteksi dini secara cepat dan akurat. Penelitian ini bertujuan untuk membangun model prediksi penyakit Tuberkulosis menggunakan algoritma Naïve Bayes dan Random Forest serta membandingkan performa kedua metode dalam proses klasifikasi penyakit TBC. Metode penelitian dilakukan melalui beberapa tahapan, yaitu pengumpulan data, preprocessing data, pembagian data latih dan data uji, pelatihan model, serta evaluasi performa model. Dataset yang digunakan terdiri dari atribut gejala dan kondisi pasien yang berkaitan dengan penyakit Tuberkulosis. Algoritma Naïve Bayes digunakan karena memiliki kemampuan klasifikasi yang sederhana dan efisien berdasarkan probabilitas, sedangkan Random Forest memanfaatkan kombinasi beberapa pohon keputusan untuk meningkatkan akurasi dan stabilitas prediksi. Evaluasi model dilakukan menggunakan metrik accuracy, precision, recall, dan F1-score. Hasil penelitian menunjukkan bahwa kedua algoritma mampu melakukan prediksi penyakit Tuberkulosis dengan baik, namun algoritma Random Forest memberikan performa yang lebih unggul dibandingkan Naïve Bayes dalam hal akurasi dan stabilitas model. Hal ini menunjukkan bahwa Random Forest lebih efektif dalam menangani data klasifikasi penyakit TBC karena mampu mengurangi risiko overfitting dan meningkatkan kemampuan generalisasi model. Berdasarkan hasil penelitian, dapat disimpulkan bahwa penerapan algoritma machine learning, khususnya Naïve Bayes dan Random Forest, dapat digunakan sebagai alternatif pendukung dalam prediksi penyakit Tuberkulosis. Penelitian ini diharapkan dapat menjadi referensi dalam pengembangan sistem prediksi penyakit berbasis kecerdasan buatan di bidang kesehatan.
Kata kunci: Tuberculosis, Prediksi Penyakit, Naïve bayes, Random Forest, Machine Learning
Abstract
Tuberculosis (TB) is one of the infectious diseases that remains a major global health problem, including in Indonesia. The high number of TB cases has increased the need for a prediction system that can support early detection quickly and accurately. This study aims to develop a Tuberculosis disease prediction model using the Naïve Bayes and Random Forest algorithms and to compare the performance of both methods in TB disease classification. The research method was conducted through several stages, including data collection, data preprocessing, data splitting into training and testing sets, model training, and model performance evaluation. The dataset used consists of several attributes related to patient symptoms and conditions associated with Tuberculosis. The Naïve Bayes algorithm was applied due to its simple and efficient probabilistic classification capability, while the Random Forest algorithm utilized an ensemble of decision trees to improve prediction accuracy and model stability. Model evaluation was carried out using accuracy, precision, recall, and F1-score metrics. The results show that both algorithms are capable of predicting Tuberculosis disease effectively; however, the Random Forest algorithm achieved better performance compared to Naïve Bayes in terms of accuracy and model stability. This indicates that the Random Forest method is more effective in handling TB disease classification data because it can reduce the risk of overfitting and improve generalization capability. Based on the findings, it can be concluded that the implementation of machine learning algorithms, especially Naïve Bayes and Random Forest, can be used as an alternative supporting tool for Tuberculosis disease prediction. This research is expected to become a reference for the development of artificial intelligence-based disease prediction systems in the healthcare field.
Keyword: Tuberculosis, Disease Prediction, Naïve Bayes, Random Forest, Machine Learning