LOKALISASI CITRA PRA-KANKER SERVIKS MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK DENGAN VISUALISASI GUIDED BACKPROPAGATION

FADHLURRAHMAN, RAYHAN DZAKI and Nurmaini, Siti (2023) LOKALISASI CITRA PRA-KANKER SERVIKS MENGGUNAKAN METODE CONVOLUTIONAL NEURAL NETWORK DENGAN VISUALISASI GUIDED BACKPROPAGATION. Undergraduate thesis, Sriwijaya Univeristy.

	Text RAMA_56201_09011281924037.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (3MB) | Request a copy
	Text RAMA_56201_09011281924037_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (9MB) | Request a copy
	Text RAMA_56201_09011281924037_0002085908_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (697kB)
	Text RAMA_56201_09011281924037_0002085908_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (182kB) | Request a copy
	Text RAMA_56201_09011281924037_0002085908_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (532kB) | Request a copy
	Text RAMA_56201_09011281924037_0002085908_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_56201_09011281924037_0002085908_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (7kB) | Request a copy
	Text RAMA_56201_09011281924037_0002085908_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (82kB) | Request a copy
	Text RAMA_56201_09011281924037_0002085908_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (881kB) | Request a copy

Abstract

The aim of this study is to perform the localization of pre-cervical cancer images using Convolutional Neural Network (CNN) method and guided backpropagation visualization. The pre-cervical cancer image data is evaluated with various CNN models that implement the second type of data augmentation. The research findings show that the CervicoNet model with epoch 200, learning rate 10-3 , and batch size 16 achieved the highest accuracy of 98.96%. The model that provided the best unseen accuracy is the CervicoNet model with epoch 150, learning rate 10-3 , and batch size 16, with an accuracy of 82.29% on previously unseen data. The best visualization results were obtained from the CervicoNet model with epoch 150, learning rate 10-3 , and batch size 16. This research is expected to contribute to the management of pre-cervical cancer through the use of deep learning techniques and guided backpropagation visualization for the localization of pre-cervical cancer images with improved accuracy

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Convolutional Neural Networks (CNN), Guided Backpropagation, Classification, Citra Pra-kanker Serviks
Subjects:	Q Science > Q Science (General) > Q1-390 Science (General) > Q223.M517 Science -- Information services. Information storage and retrieval systems --Science.
Divisions:	09-Faculty of Computer Science > 56201-Computer Systems (S1)
Depositing User:	Rayhan Dzaki Fadhlurrahman
Date Deposited:	28 Jul 2023 07:19
Last Modified:	28 Jul 2023 07:19
URI:	http://repository.unsri.ac.id/id/eprint/122544

Actions (login required)

View Item