Zahra, Nurul Kamalia (2026) Analisis Risiko Peresepan Obat Menggunakan Integrasi Algoritma C5.0 dan Association Rule Mining: Studi Kasus Puskesmas Klampis. Undergraduate thesis, UPN Veteran Jawa Timur.
|
Text (Cover)
22083010021-cover.pdf Download (1MB) | Preview |
|
|
Text (BAB 1)
22083010021-bab1.pdf Download (41kB) | Preview |
|
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (BAB 2)
22083010021-bab2.pdf Restricted to Repository staff only until 8 July 2029. Download (725kB) |
|
|
Text (BAB 3)
22083010021-bab3.pdf Restricted to Repository staff only until 8 July 2029. Download (287kB) |
|
|
Text (BAB 4)
22083010021-bab4.pdf Restricted to Repository staff only until 8 July 2029. Download (1MB) |
|
|
Text (BAB 5)
22083010021-bab5.pdf Download (19kB) | Preview |
|
|
Text (Daftar Pustaka)
22083010021-daftarpustaka.pdf Download (144kB) | Preview |
|
|
Text (Lampiran)
22083010021-lampiran.pdf Restricted to Repository staff only Download (82kB) |
Abstract
Drug prescribing is an essential process in healthcare services that may lead to various risks, including drug interactions, adverse drug reactions, and inappropriate medication use. The level of risk tends to increase as the number of prescribed medications grows and according to the characteristics of the drugs involved. Therefore, a method is needed not only to classify prescribing risk levels but also to identify patterns of relationships among drugs within prescriptions. However, previous studies have generally focused on either risk classification or drug association analysis separately. This study integrates the C5.0 algorithm and Association Rule Mining (ARM) to analyze prescription risk levels and identify drug combination patterns within each risk category. The C5.0 algorithm was employed to classify prescription data into three risk categories: high risk, moderate risk, and low risk. To simplify the model without significantly reducing its performance, the Reduced Error Pruning (REP) method was applied. Furthermore, ARM using the Apriori algorithm was utilized to discover drug association patterns within each classified risk category. The analytical results were then presented through an interactive and user-friendly Graphical User Interface (GUI). The results indicate that the C5.0 model with Reduced Error Pruning achieved an accuracy of 94.32% and a Kappa value of 0.9135. The most influential factors affecting prescription risk levels were drug category, number of prescribed drugs, and allergic reactions. In addition, each risk category exhibited distinct characteristics in terms of drug utilization and drug association patterns. The integration of C5.0 and ARM proved effective in supporting prescription risk identification and providing valuable information for prescription screening, therapy monitoring, and decision-making in healthcare facilities. Keywords: C5.0, Reduced Error Pruning, Association Rule Mining, Drug Prescribing Risk.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | T Technology > T Technology (General) > T58.6-58.62 Management Information Systems | ||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Data Science | ||||||||||||
| Depositing User: | Nurul Kamalia Zahra | ||||||||||||
| Date Deposited: | 08 Jul 2026 03:46 | ||||||||||||
| Last Modified: | 08 Jul 2026 03:46 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/54801 |
Actions (login required)
 |
View Item |