Rahmad, Said AL Hajri and -, Dr.Eng. Yusuf Sulistyo Nugroho, S.T., M.Eng. (2026) Analisis Prediksi Harga Mobil LCGC Menggunakan Multimodel Machine Learning Berdasarkan Dataset Hasil Scraping Dari Olx. Skripsi thesis, Universitas Muhammadiyah Surakarta.
![[img]](https://eprints.ums.ac.id/style/images/fileicons/application_pdf.png) |
PDF (Naskah Publikasi)
Naskah Publikasi.pdf Download (2MB) |
|
PDF (Surat Pernyataan Publikasi)
Surat Peryataan Publikasi.pdf Restricted to Repository staff only Download (845kB) |
Abstract
This study aims to analyze and compare the performance of several machine learning algorithms in predictingthe prices of used Low Cost Green Car (LCGC) vehicles using a dataset obtained from scraping the OLXplatform. The algorithms used in this study include Linear Regression, K-Nearest Neighbors (KNN), Support Vector Regression (SVR), Random Forest, and XGBoost. The dataset consisted of 14,828 used car records that underwent preprocessing stages including data cleaning, One Hot Encoding for categorical variables, andnormalization using StandardScaler. The data were then divided into training and testing sets with an 80:20ratio. Model evaluation was conducted using Mean Squared Error (MSE), coefficient of determination (R²), 5- Fold Cross-Validation, hyperparameter tuning, robustness check, and feature importance analysis. The resultsshowed that Random Forest after hyperparameter tuning became the best-performing model with an R² value of 0.802831 and an MSE value of 132,546,447,206,274.94. XGBoost achieved an R² value of 0.779038, whileLinear Regression demonstrated stable performance against noise with an R² value of 0.755665. Meanwhile, SVR produced the lowest performance with a negative R² value. In addition, the prediction model wassuccessfully implemented into a web-based system using the Flask framework to perform real-time car pricepredictions. Usability testing using the System Usability Scale (SUS) obtained an average score of 75, whichfalls into the acceptable category with a good rating. This study demonstrates that ensemble methods, particularly Random Forest and XGBoost, are capable of providing better used car price prediction performancecompared to other methods
| Item Type: | Thesis (Skripsi) |
|---|---|
| Uncontrolled Keywords: | harga mobil, machine learning, Mean Squared Error, prediksi, R-square |
| Subjects: | T Technology > T Technology (General) T Technology > Information Technology > Artificial Intelligence |
| Divisions: | Fakultas Komunikasi dan Informatika > S1 Teknik Informatika |
| Depositing User: | SAID AL HAJRI RAHMAD |
| Date Deposited: | 16 May 2026 00:44 |
| Last Modified: | 16 May 2026 00:44 |
| URI: | http://eprints.ums.ac.id/id/eprint/145376 |
Actions (login required)
 |
View Item |
