OPTIMIZATION OF CATEGORICAL BOOSTING MODEL USING OPTUNA FOR HYPERTENSION RISK PREDICTION BASED ON LIFESTYLE PATTERNS

Putri, Shalsa Adinda Dwiska (2026) OPTIMIZATION OF CATEGORICAL BOOSTING MODEL USING OPTUNA FOR HYPERTENSION RISK PREDICTION BASED ON LIFESTYLE PATTERNS. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 22081010346.-cover.pdf Download (590kB)
	Text (Bab 1) 22081010346.-bab1.pdf Download (197kB)
	Text (Bab 2) 22081010346.-bab2.pdf Restricted to Repository staff only until 21 July 2029. Download (505kB) \| Request a copy
	Text (Bab 3) 22081010346.-bab3.pdf Restricted to Repository staff only until 21 July 2029. Download (658kB) \| Request a copy
	Text (Bab 4) 22081010346.-bab4.pdf Restricted to Repository staff only until 21 July 2029. Download (1MB) \| Request a copy
	Text (Bab 5) 22081010346.-bab5.pdf Download (187kB)
	Text (Daftar Pustaka) 22081010346.-daftarpustaka.pdf Download (180kB)
	Text (Lampiran) 22081010346.-lampiran.pdf Restricted to Repository staff only Download (150kB) \| Request a copy

Abstract

Hypertension is one of the most prevalent non-communicable diseases and often remains undetected in its early stages due to the lack of noticeable symptoms. This condition can lead to serious complications, including heart disease, stroke, and kidney disorders. Therefore, a data-driven approach is needed to support faster and more accurate hypertension risk prediction. This study aims to analyze the performance of the CatBoost model in predicting hypertension risk based on lifestyle factors and to examine the effect of hyperparameter optimization using Optuna on the model's performance. This study utilized the Hypertension Risk Dataset consisting of 1,985 records. The research stages included data preprocessing, data splitting using Stratified Train-Test Split, CatBoost model development, hyperparameter optimization using Optuna based on the Tree-structured Parzen Estimator (TPE), and model evaluation using Stratified 5-Fold Cross Validation along with accuracy, precision, recall, and F1-score metrics. The results showed that, before optimization, the CatBoost model achieved an accuracy ranging from 68.84% to 98.32% and an F1-score ranging from 60.26% to 98.38%. After hyperparameter optimization using Optuna, the model performance improved, achieving an accuracy of 99.16%–99.83%, precision of 98.72%–100%, recall of 99.03%–99.68%, and F1-score of 99.20%–99.84%. These results indicate that hyperparameter optimization using Optuna effectively improved the performance of the CatBoost model for hypertension risk prediction. The optimized model was then implemented as a web-based application using Streamlit.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Muttaqin, Faisal	NIDN0030058602	faisalmuttaqin.if@upnjatim.ac.id
Thesis advisor	Anggraeny, Fetty Tri	NIDN0711028201	fettyanggraeny.if@upnjatim.ac.id

Subjects:

Q Science > QA Mathematics > QA76.87 Neural computers

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Shalsa Adinda Dwiska Putri

Date Deposited:

22 Jul 2026 06:16

Last Modified:

22 Jul 2026 07:38

URI:

https://repository.upnjatim.ac.id/id/eprint/57276