<![CDATA[The property market, particularly the housing sector, frequently experiences price fluctuations influenced by various complex factors such as location, building size, number of rooms, and economic and demographic conditions. This price uncertainty often makes it difficult for sellers to set competitive selling prices and makes it difficult for buyers to estimate the fair value of a residence. This study aims to develop a house price prediction model using a Machine Learning algorithm, namely Multiple Linear Regression. The research data was obtained from the Kaggle public repository (USA Housing Dataset), which contains 5,000 transaction data with independent variables being the average regional income, building age, number of rooms, number of bedrooms, and regional population. The research methodology follows the stages of Knowledge Discovery in Database (KDD), starting from data collection, cleaning, transformation, modeling, and evaluation. The model was built using the Python programming language with the help of the Scikit-Learn library. The test results show that linear regression is able to provide house price predictions with very good performance, indicated by an R-Squared (R²) value of 0.91, a Mean Absolute Error (MAE) of 81,250, and a Root Mean Squared Error (RMSE) of 100,500. These findings indicate that demographic characteristics and physical attributes of buildings have a strong linear relationship to house selling prices. Keywords: Data Mining, Machine Learning, House Price Prediction, Python, Linear Regression. Pasar properti, khususnya sektor perumahan, sering mengalami perubahan harga yang dipengaruhi oleh berbagai faktor kompleks seperti letak wilayah, luas bangunan, jumlah ruangan, serta kondisi ekonomi dan demografi. Ketidakpastian harga tersebut kerap menyulitkan penjual dalam menetapkan harga jual yang kompetitif dan membuat pembeli kesulitan memperkirakan nilai wajar suatu hunian. Penelitian ini bertujuan mengembangkan model prediksi harga rumah dengan memanfaatkan algoritma Machine Learning, yaitu Multiple Linear Regression. Data penelitian diperoleh dari repositori publik Kaggle (USA Housing Dataset) yang berisi 5.000 data transaksi dengan variabel bebas berupa rata-rata pendapatan kawasan, usia bangunan, jumlah ruangan, jumlah kamar tidur, dan populasi wilayah. Metodologi penelitian mengikuti tahapan Knowledge Discovery in Database (KDD), mulai dari pengumpulan data, pembersihan, transformasi, pemodelan, hingga evaluasi. Model dibangun menggunakan bahasa pemrograman Python dengan bantuan pustaka Scikit-Learn. Hasil pengujian menunjukkan bahwa regresi linear mampu memberikan prediksi harga rumah dengan performa yang sangat baik, ditunjukkan oleh nilai R-Squared (R²) sebesar 0.91, Mean Absolute Error (MAE) sebesar 81.250, dan Root Mean Squared Error (RMSE) sebesar 100.500. Temuan ini mengindikasikan bahwa karakteristik demografi dan atribut fisik bangunan memiliki hubungan linear yang kuat terhadap harga jual rumah. Kata Kunci: Data Mining, Machine Learning, Prediksi Harga Rumah, Python, Regresi Linier.]]>
