Analisis Model Klasifikasi untuk Prediksi Penyakit Liver Menggunakan Regresi Logistik dan Random Forest

Abstract

Liver disease is a serious healthcondition that requires early detection to prevent further complications. This study aims to evaluate the performance of classification models in predicting liver disease using Logistic Regression and Random Forest algorithms. The dataset was obtained from the Kaggle platform and includes clinical variabels such as age, gender, bilirubin levels, liver enzymes, protein levels, and albumin. The research stages consist of data preprocessing, exploratory data analysis (EDA), data splitting into training and testing sets, and model development using both algorithms. Model performance was evaluated using accuracy, precision, recall, and F1-Score based on the confusion matrix. The results indicate that Logistic Regression achieves a very high recall value of 0.99, making it effective in detecting positive liver disease cases. Meanwhile, Random Forest demonstrates a more stable performance with an accuracy of 0.75 and precision of 0.76. Both models obtain the same F1-Score of 0.84, indicating a balanced performance between precision and recall. Overall, both algorithms provide reliable prediction results, with Logistic Regression being more suitable for early detection and Random Forest offering more stable and balanced predictions.