Penerapan Model XGB-ILSO-1DCNN untuk Klasifikasi Anemia Berdasarkan Data CBC

Febrianti, Raissa Atha (2025) Penerapan Model XGB-ILSO-1DCNN untuk Klasifikasi Anemia Berdasarkan Data CBC. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21081010244-cover.pdf Download (1MB)
	Text (Bab 1) 21081010244.-bab1.pdf Download (23kB)
	Text (Bab 2) 21081010244.-bab2.pdf Restricted to Repository staff only until 17 December 2028. Download (481kB)
	Text (Bab 3) 21081010244.-bab3.pdf Restricted to Repository staff only until 17 December 2028. Download (512kB)
	Text (Bab 4) 21081010244.-bab4.pdf Restricted to Repository staff only until 17 December 2028. Download (1MB)
	Text (Bab 5) 21081010244.-bab5.pdf Download (11kB)
	Text (Daftar pustaka) 21081010244.-daftarpustaka.pdf Download (139kB)
	Text (Lampiran) 21081010244.-lampiran.pdf Restricted to Repository staff only until 17 December 2028. Download (150kB)

Abstract

This study aims to classify types of anemia based on Complete Blood Count (CBC) data by applying the XGBoost–Improved Lion Swarm Optimization–One Dimensional Convolutional Neural Network (XGB-ILSO-1DCNN) model. The research method consists of eight main stages, namely collecting the CBC dataset from Mendeley Data, preprocessing (data cleaning, categorical transformation, feature standardization, and data division), feature augmentation using XGBoost, data balancing with the Synthetic Minority Oversampling Technique (SMOTE), data transformation to 1D-CNN format, hyperparameter optimization using ILSO, model training, and performance evaluation using a Confusion Matrix. The XGBoost model was used to generate additional feature representations that enriched the input data, while 1D-CNN served as the main classification model to recognize patterns between hematological features. The optimization process was carried out using the ILSO algorithm to adjust important parameters such as learning rate, dropout, batch size, and epoch in order to obtain the best model configuration. The results show that the application of the ILSO algorithm significantly improves model performance compared to models without optimization. The XGB-ILSO-1DCNN model produced the highest accuracy of 98.63%, with a precision value of 99.32%, a recall of 97.50%, and an F1-score of 98.34%. Based on these results, the application of the XGB-ILSO-1DCNN hybrid approach can improve the accuracy, stability, and generalization ability of the model, so that it can be used as a clinical decision support system for automatic and efficient anemia diagnosis.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Sari, Anggraini Puspita	NIDN0716088605	anggraini.puspita.if@upnjatim.ac.id
Thesis advisor	Nurlaili, Afina Lina	NIDN0013129303	afina.lina.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Atha Raissa

Date Deposited:

17 Dec 2025 05:09

Last Modified:

17 Dec 2025 05:09

URI:

https://repository.upnjatim.ac.id/id/eprint/48212