Bulletin of Artificial Intelligence

Klasifikasi Penyakit Ginjal Kronis pada Data Tidak Seimbang Menggunakan K-Nearest Neighbor Berbasis Seleksi Fitur Mutual Information dan GridSearchCV

2026-07-15T14:14:45+00:00

Chronic Kidney Disease (CKD) is a progressive disease characterized by a gradual decline in kidney function and requires early detection to reduce the risk of severe complications. Machine learning has been widely applied to support CKD classification based on clinical attributes; however, medical datasets often contain missing values, a combination of numerical and categorical features, and class imbalance. This study aims to evaluate the performance of the K-Nearest Neighbor (KNN) algorithm for CKD classification using Mutual Information feature selection and GridSearchCV. The dataset consisted of 400 samples, including 250 CKD cases and 150 non-CKD cases. The proposed methodology included data cleaning, missing value imputation, categorical feature encoding, numerical feature normalization using MinMaxScaler, feature selection using SelectKBest with Mutual Information, and hyperparameter tuning using GridSearchCV. Model performance was evaluated using hold-out testing and 10-fold cross-validation. The hold-out evaluation showed that the KNN model with GridSearchCV achieved 100.00% accuracy, precision, recall, F1-score, and AUC on the test set. To ensure that this result was not dependent on a single train-test split, additional evaluation was conducted using 10-fold cross-validation. The cross-validation results yielded an average accuracy of 99.25% for the KNN model with GridSearchCV, indicating consistent performance across different data partitions. Meanwhile, the KNN model with Mutual Information feature selection and GridSearchCV achieved 98.75% accuracy, 100.00% recall, and a 99.01% F1-score, demonstrating competitive performance while using a more compact feature subset. The findings indicate that the application of GridSearchCV improved the performance of the KNN model on the dataset used, while Mutual Information contributed to selecting relevant features, enabling the model to maintain strong classification performance with a reduced number of features

Prediksi Magnitudo Gempa Indonesia Menggunakan Machine Learning Berbasis Data Seismik

2026-07-15T16:26:22+00:00

Indonesia is one of the world's most seismically active regions, making earthquake magnitude prediction an important component of disaster mitigation. This study compares several machine learning regression algorithms for predicting earthquake magnitude using Indonesian seismic data collected from the United States Geological Survey (USGS) during 2016–2023. The dataset contains 17,331 earthquake records consisting of spatial, measurement-quality, and temporal attributes. The research process included data preprocessing, feature engineering, model training, and evaluation using the Coefficient of Determination (R²), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE). Four regression algorithms were evaluated: Linear Regression, Support Vector Regression, Random Forest Regression, and Gradient Boosting Regression. Experimental results show that Random Forest Regression achieved the best performance with an R² of 0.767, MAE of 0.123, and RMSE of 0.179, followed by Gradient Boosting Regression with an R² of 0.746. Feature importance analysis indicates that magError, depthError, depth, and gap are the most influential variables in earthquake magnitude prediction. These findings demonstrate that ensemble learning provides accurate prediction performance while offering meaningful interpretation of influential seismic parameters for disaster mitigation systems

Analisis Sentimen Pengaruh Media Sosial Terhadap Keputusan Pembelian Konsumen Menggunakan Metode K-Nearest Neighbors (K-NN)

2026-07-29T04:59:14+00:00

Perkembangan media sosial telah mengubah cara konsumen berinteraksi dan mengambil keputusan pembelian. Media sosial menjadi platform utama bagi konsumen untuk berbagi pengalaman, memberikan ulasan, dan mendiskusikan produk atau layanan. Penelitian ini bertujuan untuk menganalisis sentimen konsumen terhadap ulasan produk dimedia sosial serta dampaknya pada keputusan pembelian menggunakan metode K-NN. Data yang digunakan berupa ulasan konsumen yang dikumpulkaan pada platform media sosial. Data tersebut diproses melalui tahapan preprocessing, seperti tokenisasi, penghapusan kata tidak penting, dan stemming, sebelum diklasifikasikan menjadi sentimen positif dan negatif. K-NN dipilih karena keandalannya dalam klasifikasi berbasis jarak dan kemampuannya menangani datasetdengan pola distribusi yang kompleks. Hasil penelitian menunjukkan bahwa metode K-NN mampu mengklasifikasikan sentimen konsumen dengan akurasi 100%. Hasil analisis menunjukkan bahwa sentimen positif berkonstribusi signifikan dalam mempengaruhi keputusan keputusan pembelian, sedangkan sentimen negatif cenderung mengurangi minat konsumen. Hasil yang diperoleh dari tahap modelling dengan menggunakan metode K-NN dan perbandingan 80:20 untuk data training dan testing, maka nilai akurasi yang dihasilkan sebesar 90%, precision 90%, recall 100%. Evaluasi model dilakukan menggunakan matrik akurasi, presisi dan recall, dengan hasil yang menunjukkan kinerja K-NN cukup baik dalam mengklasifikasikan sentimen konsumen. Penelitian ini memberikan wawasan tentang pentingnya memahami sentimen konsumen di media sosial untuk strategi pemasaran yang lebih efektif, serta menunjukkan potensi penerapan machine learning, khususnya K-NN dalam menganalisis data tekstual untuk mendukung pengambilan keputusan bisnis.