HARTONO, ARDAN DWI (2025) THE IMPLEMENTATION OF REINFORCEMENT LEARNING IN A MUSIC RECOMMENDATION SYSTEM APPLICATION BASED ON USER BEHAVIOR. Tugas Akhir thesis, Informatics.
![[img]](http://eprints.uty.ac.id/style/images/fileicons/text.png) |
Text
5210411077_Ardan Dwi Hartono_Abstrak.pdf Download (63kB) |
Abstract
Music serves not only as a form of entertainment but also as a medium for emotional expression through sound and rhythm. Often consumed sequentially within listening sessions or playlists, music requires personalized recommendations to enhance user experience. This study successfully develops a user behavior–based music recommendation system using Collaborative Filtering and Reinforcement Learning methods. The system dynamically adapts to user interactions in real time. Two exploration strategies—Thompson Sampling and Upper Confidence Bound (UCB)—are implemented to optimize recommendation accuracy. Based on manual evaluations using 15 user action data points, Thompson Sampling selected “Bohemian Rhapsody” by Queen with a score of 0.71, though it demonstrated instability and a tendency to explore low-reward items. In contrast, the UCB approach selected “I'm Yours” by Jason Mraz with a higher and more stable score of 2.177, attributed to its explicit calculation of reward averages and confidence intervals, resulting in more controlled exploration. Keywords: Music, Recommendation System, Reinforcement Learning, Thompson Sampling, Upper Confidence Bound.
| Item Type: | Thesis (Skripsi, Tugas Akhir or Kerja Praktek) (Tugas Akhir) |
|---|---|
| Subjects: | T Technology > T Technology (General) |
| Divisions: | Fakultas Sains Dan Teknologi > S1 Informatika |
| Depositing User: | Kaprodi S1 Informatika UTY |
| Date Deposited: | 19 Jul 2025 02:13 |
| Last Modified: | 19 Jul 2025 02:13 |
| URI: | http://eprints.uty.ac.id/id/eprint/18279 |
Actions (login required)
 |
View Item |
