Perbandingan Kinerja Metode XGBoost dan K-NN Enhanced dengan PCA dalam Prediksi Tingkat Keparahan Pasien Hemodialisis

Wafa, Mochammad Thoriq (2025) Perbandingan Kinerja Metode XGBoost dan K-NN Enhanced dengan PCA dalam Prediksi Tingkat Keparahan Pasien Hemodialisis. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) Cover.pdf Download (819kB)
	Text (Bab 1) Bab 1.pdf Download (195kB)
	Text (Bab 2) Bab 2.pdf Restricted to Repository staff only until 28 November 2028. Download (503kB)
	Text (Bab 3) Bab 3.pdf Restricted to Repository staff only until 28 November 2028. Download (586kB)
	Text (Bab 4) Bab 4.pdf Restricted to Repository staff only until 28 November 2028. Download (1MB)
	Text (Bab 5) Bab 5.pdf Download (173kB)
	Text (Daftar Pustaka) Daftar Pustaka.pdf Download (161kB)
	Text (Lampiran) Lampiran.pdf Restricted to Repository staff only until 28 November 2028. Download (594kB)

Abstract

This study compares the performance of Extreme Gradient Boosting (XGBoost) and K-Nearest Neighbors Enhanced (K-NN Enhanced), each evaluated with and without Principal Component Analysis (PCA), for classifying the severity level of hemodialysis patients. The dataset was constructed from selected clinical parameters and preprocessed through column alignment, missing-value handling, categorical encoding, MinMax normalization, and class balancing using Random OverSampling. The data were then split stratified into 80 percent training, 10 percent validation, and 10 percent testing subsets. Hyperparameters were optimized using ten-fold GridSearchCV, and model evaluation on the test set employed accuracy, precision, recall, macro-averaged F1-score, and confusion matrix analysis. The results show that XGBoost without PCA achieved the best performance, with an accuracy of 91.65 percent, precision 0.92, recall 0.92, and F1-score 0.92. PCA improved the K-NN Enhanced model from 82.84 percent to 84.12 percent but slightly reduced XGBoost performance from 91.65 percent to 90.47 percent. These findings indicate that dimensionality reduction should be aligned with algorithm characteristics and that XGBoost is the most reliable model for predicting hemodialysis severity in this dataset.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Puspaningrum, Eva Yulia	NIDN0005078908	evapuspaningrum.if@upnjatim.ac.id
Thesis advisor	Rahajoe, Rr. Ani Dijah	NIDN0012057301	anidijah.if@upnjatim.ac.id

Subjects:

R Medicine > RA Public aspects of medicine
R Medicine > RA Public aspects of medicine > RA0421 Public health. Hygiene. Preventive Medicine
T Technology > TK Electrical engineering. Electronics Nuclear engineering
T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105 Computer Network

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Mr. Thoriq Wafa

Date Deposited:

28 Nov 2025 08:21

Last Modified:

28 Nov 2025 08:47

URI:

https://repository.upnjatim.ac.id/id/eprint/47057