ANALISIS PERFORMANSI MODEL VGG-16 DAN VGG-16-ELM BERDASARKAN VARIASI UKURAN INPUT CITRA DAN BALANCING DATA UNTUK KLASIFIKASI PNEUMONIA

Zein, Mohammad Agil Rofiqul (2025) ANALISIS PERFORMANSI MODEL VGG-16 DAN VGG-16-ELM BERDASARKAN VARIASI UKURAN INPUT CITRA DAN BALANCING DATA UNTUK KLASIFIKASI PNEUMONIA. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21081010117-cover.pdf Download (2MB)
	Text (Bab 1) 21081010117-bab 1.pdf Download (273kB)
	Text (Bab 2) 21081010117-bab 2.pdf Restricted to Repository staff only until 20 June 2027. Download (727kB)
	Text (Bab 3) 21081010117-bab 3.pdf Restricted to Repository staff only until 20 June 2027. Download (500kB)
	Text (Bab 4) 21081010117-bab 4.pdf Restricted to Repository staff only until 20 June 2027. Download (3MB)
	Text (Bab 5) 21081010117-bab 5.pdf Download (262kB)
	Text (Daftar pustaka) 21081010117-daftar pustaka.pdf Download (228kB)

Abstract

Pneumonia is a lung disease that can be detected through chest X-ray images. This study aims to improve the accuracy and efficiency of automatic diagnosis by analyzing the performance of two deep learning models, namely VGG 16 and the combination of VGG-16 with Extreme Learning Machine (ELM), in pneumonia classification. The focus of the research is on analyzing the effect of variations in input image sizes (150×150, 200×200, 224×224, 256×256, and 300×300 pixels) and applying data balancing techniques using Random Over Sampling (ROS). The dataset used consists of 5,856 X-ray images divided into two classes: NORMAL and PNEUMONIA. The preprocessing stages include resizing, normalization, data splitting, and augmentation. Performance evaluation is conducted using accuracy, precision, recall, and F1-score metrics. The research results show that an input size of 200×200 yields the best results across all scenarios. The VGG-16 model without ROS achieved the highest accuracy of 96.59% and an F1-score of 97.69%. Meanwhile, the VGG-16-ELM combination demonstrated improved performance when ROS was applied. These findings emphasize that model architecture, data balancing techniques, and image input size play a crucial role in classification accuracy and can support the development of artificial intelligence-based pneumonia diagnostic systems. Keywords : Pneumonia, VGG-16, Extreme Learning Machine, Random Over Sampling, Image Classification, Deep Learning.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Rahmat, Basuki	NIDN0023076907	basukirahmat.if@upnjatim.ac.id
Thesis advisor	Puspaningrum, Eva Yulia	NIDN0005078908	evapuspaningrum.if@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Informatics

Depositing User:

Mr Agil Zein

Date Deposited:

20 Jun 2025 02:41

Last Modified:

20 Jun 2025 02:41

URI:

https://repository.upnjatim.ac.id/id/eprint/38753