Prasetyo, Bagus Rizky (2024) Komparasi Performa Model Berbasis Algoritma Random Forest dan LightGBM dalam Melakukan Klasifikasi Diabetes Melitus Gestasional. Undergraduate thesis, UPN Veteran Jawa Timur.
![[img]](https://repository.upnjatim.ac.id/style/images/fileicons/text.png) |
Text (Cover)
19082010062-cover.pdf Download (732kB) |
|
Text (Bab 1)
19082010062-bab1.pdf Download (441kB) |
|
Text (Bab 2)
19082010062-bab2.pdf Restricted to Repository staff only until 12 July 2026. Download (1MB) |
|
Text (Bab 3)
19082010062-bab3.pdf Restricted to Repository staff only until 12 July 2026. Download (1MB) |
|
Text (Bab 4)
19082010062-bab4.pdf Restricted to Repository staff only until 12 July 2026. Download (2MB) |
|
Text (BAB 5)
19082010062-bab5.pdf Download (403kB) |
|
Text (Daftar Pustaka)
19082010062-daftarpustaka.pdf Download (319kB) |
|
Text (Lampiran)
19082010062-lampiran.pdf Restricted to Repository staff only until 12 July 2026. Download (542kB) |
Abstract
Gestational diabetes mellitus (GDM) is a condition characterized by an increase in blood sugar levels in a mother during pregnancy. Mothers indicated to have gestational diabetes mellitus are at risk of several serious complications if not properly managed. Early detection through the use of medical record data is carried out as a preventive measure to avoid complications in the future. The performance of the random forest and LightGBM algorithms in predicting medical record data will be compared to determine which is superior in terms of classification success and CPU efficiency. The CRISP-DM framework is utilized in the model development. Data is obtained from the obstetrics clinic of the Islamic Hospital Surabaya Jemursari, consisting of 270 rows and 20 columns. After the data preparation phase, 267 rows and 11 columns remain for model creation. Modeling is conducted in 18 scenarios. The best scenario is the default parameter random forest with class imbalance handling using ADASYN on a 70:30 data split. This model achieved an accuracy of 88%, precision of 27%, recall of 100%, and F1 score of 43%. The model is implemented on the CheckDMG website application, built using Flask. Validation testing results on the website achieved an accuracy of 85%.
| Item Type: | Thesis (Undergraduate) | ||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Contributors: |
|
||||||||||||
| Subjects: | Q Science > Q Science (General) Q Science > QA Mathematics > QA76 Computer software Q Science > QA Mathematics > QA76.6 Computer Programming T Technology > T Technology (General) |
||||||||||||
| Divisions: | Faculty of Computer Science > Departemen of Information Systems | ||||||||||||
| Depositing User: | Bagus Rizky Prasetyo | ||||||||||||
| Date Deposited: | 12 Jul 2024 09:00 | ||||||||||||
| Last Modified: | 12 Jul 2024 09:00 | ||||||||||||
| URI: | https://repository.upnjatim.ac.id/id/eprint/25898 |
Actions (login required)
 |
View Item |