Penggunaan Algoritma Decision Tree Untuk Mengukur Tingkat Kepuasan Pengguna Aplikasi KAI Access
| Penulis |
|---|
| Dosen: Mahasiswa: |
| Tanggal Terbit |
| 01 April 2024 |
| Kategori |
| Jurnal Nasional Terakreditasi [Lainnya] |
| Penerbit |
| Kohesi Jurnal Sains dan Teknologi |
| Kota / Negara |
| Kohesi / Indonesia |
| Volume |
| 3/1 |
| Halaman |
| 51-60 |
| E-ISSN |
| 3025-1311 |
| DOI |
| https://doi.org/10.3785/kohesi.v3i1 |
| URL |
| https://ejournal.warunayama.org/index.php/kohesi/article/view/2875 |
| Abstrak |
| Transportation is crucial in a region's economy and society's movement. Trains (KA) are currently a popular mode of transportation due to their ease of use, security and dependability. The only State-Owned Enterprise (BUMN) that provides rail transportation services is the Indonesian Railroad Company (KAI). The effort to meet customer needs and wants and the precision of delivery in line with customer expectations are all aspects of service quality. Customers are more likely to be satisfied with the company if the quality of its service is higher. To improve accuracy, the Optimize Selection method and Split Validation are used. These research models use the Decision Tree algorithm with a dataset of user satisfaction from the KAI Access application obtained from questionnaires. Training data and testing data make up the research dataset. The cross-validation and split-validation operators will be used to divide the data. The validation results using 10-fold Validation on the Decision Tree algorithm show the best performance compared to the other two algorithms, namely Naive Bayes and K-Nearest Neighbor (K-NN). The Decision Tree algorithm yields an accuracy of 98.96% and an AUC value of 0.800. By optimizing feature selection and validation using Split Validation with a ratio of 0.8, an accuracy of 100% and an AUC value of 1.000 are achieved. |
