OPTIMASI ALGORITMA DEEP NEURAL NETWORK MENGGUNAKAN INISIALISASI WEIGHT KAIMING HE UNTUK KLASIFIKASI PENYAKIT SERANGAN JANTUNG

ANDIANI, LIA and Sukemi, Sukemi and Rini, Dian Palupi (2020) OPTIMASI ALGORITMA DEEP NEURAL NETWORK MENGGUNAKAN INISIALISASI WEIGHT KAIMING HE UNTUK KLASIFIKASI PENYAKIT SERANGAN JANTUNG. Master thesis, Sriwijaya University.

	Text RAMA_55101_09042681822010.pdf - Accepted Version Restricted to Registered users only Available under License Creative Commons Public Domain Dedication. Download (2MB) | Request a copy
Preview	Text RAMA_55101_09042681822010_0003126604_0023027804_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB) | Preview
	Text RAMA_55101_09042681822010_0003126604_0023027804_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (590kB) | Request a copy
	Text RAMA_55101_09042681822010_0003126604_0023027804_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (696kB) | Request a copy
	Text RAMA_55101_09042681822010_0003126604_0023027804_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_55101_09042681822010_0003126604_0023027804_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (258kB) | Request a copy
	Text RAMA_55101_09042681822010_0003126604_0023027804_06_ref.pdf - Published Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (469kB) | Request a copy
	Text RAMA_55101_09042681822010_0003126604_0023027804_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (1MB) | Request a copy
	Text RAMA_55101_09042681822010_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (9MB) | Request a copy

Abstract

The Disease of the heart or cardiovascular organs is the number one cause of death in 17.7 million deaths in the world. Coronary Heart Disease (CHD) is increasing every year with a significant number of deaths. In Indonesia, the highest percentage of cardiovascular disease is coronary heart disease, coronary heart disease, which is 1.5 percent. The aim of this study is to minimize the expert diagnosis time and increase the accuracy of diagnosis. DNN is a neural network-based algorithm that can be used for decisions that have more than one hidden neural layer. This algorithm is the development of intelligence, namely the Artificial Neural Network (ANN) algorithm. To achieve high accuracy in this algorithm, the amount of data needs to be trained first. The accuracy of the system to be developed can be improved by adding a Kaiming He weight initialization optimization technique to the DNN structure. Therefore, this study proposes that DNN be optimized with a Kaiming He weight initialization technique so that it can increase the accuracy, sensitivity, and specificity values, and can overcome weaknesses in large data variants between classes. This is evidenced by the results of the accuracy performance of 98.73%, 99.21% precision, 99.11% sensitivity, and 98.36% specificity.

Item Type:	Thesis (Master)
Uncontrolled Keywords:	CHD, Kaiming He, DNN, Accuracy, Sensitivity, Specificity
Subjects:	T Technology > T Technology (General) > T58.5-58.64 Information technology > T58.5 General works Management information systems Cf. HD30.213 Industrial management Cf. HF5549.5.C6+ Communication in personnel management Cf. TS158.6 Automatic data collection systems (Production control)
Divisions:	09-Faculty of Computer Science > 55101-Informatics (S2)
Depositing User:	Users 10905 not found.
Date Deposited:	16 Feb 2021 08:04
Last Modified:	16 Feb 2021 08:04
URI:	http://repository.unsri.ac.id/id/eprint/42988

Actions (login required)

View Item