Mohamad Rizal Hanafi, - (2024) RANCANG BANGUN APLIKASI KLASIFIKASI LEVEL TRIASE PADA PASIEN IGD MENGGUNAKAN ARSITEKTUR CNN SATU DIMENSI. S1 thesis, Universitas Pendidikan Indonesia.
![]() |
Text
S_TEKOM_2005678_Title.pdf Download (485kB) |
![]() |
Text
S_TEKOM_2005678_Chapter 1.pdf Download (241kB) |
![]() |
Text
S_TEKOM_2005678_Chapter 2.pdf Restricted to Staf Perpustakaan Download (672kB) |
![]() |
Text
S_TEKOM_2005678_Chapter 3.pdf Download (900kB) |
![]() |
Text
S_TEKOM_2005678_Chapter 4.pdf Restricted to Staf Perpustakaan Download (1MB) |
![]() |
Text
S_TEKOM_2005678_Chapter 5.pdf Download (217kB) |
![]() |
Text
S_TEKOM_2005678_Appendix.pdf Restricted to Staf Perpustakaan Download (781kB) |
Abstract
Triase merupakan solusi dari ketidakseimbangan tenaga kesehatan yang tersedia di IGD dengan jumlah pasien yang berkunjung. Permasalahan tersebut disebut dengan overcrowded, yaitu isu umum yang terjadi di berbagai negara termasuk di Indonesia. Meskipun demikian, dengan adanya triase tidak menyelesaikan masalah tersebut. Berdasarkan temuan pada penelitian terdahulu, masalah tersebut telah dicoba untuk diselesaikan. Namun, terdapat kekurangan dari penelitian-penelitian terdahulu yaitu model klasifikasi level triase disederhanakan menjadi dua atau tiga kelas. Ini membuat klasifikasi level triase tidak sesuai dengan standar ESI, yaitu triase lima level. Oleh karena itu, penelitian ini bertujuan untuk mengimplementasikan model klasifikasi level triase dengan arsitektur CNN satu dimensi kedalam aplikasi berbasis web untuk membandingkan dan mengevaluasi kinerjanya dengan model lain seperti neural network, XGBoost, dan logistic regression. Metode yang digunakan dalam penelitian ini adalah design and development dengan pendekatan AI cycle project. Hasil penelitian menunjukkan akurasi dari arsitektur CNN satu dimensi bernilai 81% dengan recall, precision, dan f1-score memiliki nilai 0.81. Model dengan arsitektur CNN satu dimensi ini memiliki kinerja paling baik diikuti dengan neural network (akurasi 78%; precision 0,78; recall 0,77; dan f1-score 0,78) XGBoost (Akurasi 75%; precision 0,76; recall 0,75; f1-score 0.75), dan logistic regression (Akurasi 70%; precision 0,70; recall 0,70; dan f1-score 0,70). Selain itu, pengujian black-box menunjukkan seluruh fitur yang telah dirancang pada aplikasi dapat berfungsi dengan baik sesuai dengan yang diharapkan. --------- Triage is a solution to the imbalance of health workers available in the emergency room with the number of visiting patients. This problem is called overcrowded, which is a common issue in many countries including Indonesia. However, triage does not solve the problem. Based on the findings of previous studies, the problem has been tried to be solved. However, a shortcoming of previous studies is that the triage level classification model is simplified to two or three classes. This makes the triage level classification incompatible with the ESI standard, which is five-level triage. Therefore, this study aims to implement a triage level classification model with one-dimensional CNN architecture into a web-based application to compare and evaluate its performance with other models such as neural network, XGBoost, and logistic regression. The method used in this research is design and development with the AI cycle project approach. The results showed that the accuracy of the one-dimensional CNN architecture was 81% with recall, precision, and f1-score having a value of 0.81. This model with one-dimensional CNN architecture has the best performance followed by neural network (78% accuracy; precision 0.78; recall 0.77; and f1-score 0.78) XGBoost (75% accuracy; precision 0.76; recall 0.75; f1-score 0.75), and logistic regression (70% accuracy; precision 0.70; recall 0.70; and f1-score 0.70). In addition, black-box testing shows that all features that have been designed in the application can function properly as expected.
Item Type: | Thesis (S1) |
---|---|
Additional Information: | https://scholar.google.com/citations?user=HTIcPkUAAAAJ&hl=id Sinta ID : 6745899 Sinta ID : 6680849 |
Uncontrolled Keywords: | Klasifikasi Level Triase, Deep Learning, CNN satu dimensi, Pengujian Black-box, Triage Level Classification, Deep Learning, one-dimensional CNN, Black-box Testing |
Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software T Technology > T Technology (General) |
Divisions: | UPI Kampus cibiru > S1 Teknik Komputer |
Depositing User: | Mohamad Rizal Hanafi |
Date Deposited: | 23 Aug 2024 03:22 |
Last Modified: | 23 Aug 2024 03:22 |
URI: | http://repository.upi.edu/id/eprint/120201 |
Actions (login required)
![]() |
View Item |