PENERAPAN JARINGAN SYARAF TIRUAN MENGGUNAKAN METODE BACKPROPAGATION UNTUK MENDETEKSI PENYAKIT BRONKITIS 2017

PUTRI, ANDRIANI and Irmawan, Irmawan (2017) PENERAPAN JARINGAN SYARAF TIRUAN MENGGUNAKAN METODE BACKPROPAGATION UNTUK MENDETEKSI PENYAKIT BRONKITIS 2017. Undergraduate thesis, Sriwijaya University.

	Text RAMA_20201_03121404064.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (21MB) | Request a copy
	Text RAMA_20201_03121404064_TURNITIN.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (8MB) | Request a copy
Preview	Text RAMA_20201_03121404064_0017097405_01_front_ref.pdf - Accepted Version Available under License Creative Commons Public Domain Dedication. Download (1MB) | Preview
	Text RAMA_20201_03121404064_0017097405_02.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (363kB) | Request a copy
	Text RAMA_20201_03121404064_0017097405_03.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (129kB) | Request a copy
	Text RAMA_20201_03121404064_0017097405_04.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (982kB) | Request a copy
	Text RAMA_20201_03121404064_0017097405_05.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (156kB) | Request a copy
	Text RAMA_20201_03121404064_0017097405_06_ref.pdf - Bibliography Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (251kB) | Request a copy
	Text RAMA_20201_03121404064_0017097405_07_lamp.pdf - Accepted Version Restricted to Repository staff only Available under License Creative Commons Public Domain Dedication. Download (18MB) | Request a copy

Abstract

Based on research that has been done at lungs hospital in Palembang about bronchitis, it is known that there are 10 disease which is high suffered by outpatient in hospital. During the period 2016 bronchitis disease has 2214 patients in 2016. Use of Artificial Neural Networks here using Backpropaagtion algorithm on Matlab toolbox which is used as a learning algorithm in determining whether or not the patient suspected bronchitis. After determining the general characteristics and record the results of the patient's medical records. A total of 80 patient data for train data and 20 patient data for test data. Determination of network architecture in this research using trial and error repeatedly to get the smallest Mean Square Error (MSE) value of all existing tests and the fastest time compared to other experiments. To test the performance of the program, the experiment was conducted where in the experimental trial of making bronchitis disease detection system using bronchitis patient data, the best result after 20 different variations using confusion matrix accuracy was obtained with 100% accuracy level with the smallest MSE 0.009003 on the 12th data. The best results of the network training using 100 hidden layer neurons, learning rate of 0.35, target error is 0.001, epoch 1000 and trainign time for 4 minutes 23 seconds. In the second experiment, the addition of patients with ISPA disease. The result of the training was obtained after 20 variations using confusion matrix with the smallest MSE that is 0,00717 on the 19th data. In the 19th data has a high level of accuracy is 99.5% and the fastest training time data is 2 seconds. To get the best result is using the learning rate of 0.8, 70 hidden hidden neurons, and epoch 1000, and target error of 0.001. Based on the experiments conducted on the results of both trials the addition of ISPA disease although decreased accuracy level but the proof of suspected or not bronchitis patients is still well detected, and can be seen the comparison between them starting from the learning rate (α), hidden layer neurons, epoch, and target error. The best results are taken for the accuracy of the detection system.

Item Type:	Thesis (Undergraduate)
Uncontrolled Keywords:	Penerapan Jaringan Syaraf Tiruan, Metode Backpropagation
Subjects:	R Medicine > RV Botanic, Thomsonian, and eclectic medicine T Technology > TT Handicrafts Arts and crafts
Divisions:	03-Faculty of Engineering > 20201-Electrical Engineering (S1)
Depositing User:	Users 3043 not found.
Date Deposited:	05 Dec 2019 04:48
Last Modified:	05 Dec 2019 04:48
URI:	http://repository.unsri.ac.id/id/eprint/20026

Actions (login required)

View Item