PERBANDINGAN JARINGAN SYARAF TIRUAN DAN NAIVE BAYES DALAM DETEKSI SESEORANG TERKENA PENYAKIT STROKE

I. Rohmana(1), R. Arifudin(2),


(1) Gedung D7 Lantai 1, Kampus Unnes Sekaran, Gunungpati, Semarang, 50229
(2) Gedung D7 Lantai 1, Kampus Unnes Sekaran, Gunungpati, Semarang, 50229

Abstract

Tujuan penelitian ini adalah membuat aplikasi Jaringan Syaraf Tiruan dan Naive Bayes untuk memprediksi penyakit stroke dan membandingkan tingkat akuratan dari kedua metode yang digunakan. Sebuah aplikasi software MATLAB diciptakan untuk mendeteksi seseorang Suspect stroke.  Metode yang baik dalam mesin pembelajaran berdasarkan data training adalah Jaringan Syaraf Tiruan dan Naive Bayes, variabel data faktor gejala penyakit stroke digunakan sebagai data training dalam proses pembelajaran dari sistem yang dibuat menentukan prediksi penyakit stroke. Dari 120 data percobaan yang dilakukan, akan dihitung akurasi hasil kerja sistem yang dibagi menjadi data pelatihan dan data pengujian. Diperoleh persentase hasil kerja sistem yaitu Jaringan Syaraf Tiruan sebesar 71,11 persen, sedangkan Naive Bayes sebesar 80,55 persen. Naive Bayes lebih akurat daripada Jaringan Syaraf Tiruan dalam hal pengambilan keputusan data baru namun Jaringan Syaraf Tiruan memiliki teknik yang lebih bagus dibandingkan dengan Naive Bayes. Jaringan Syaraf Tiruan mempunyai karakteristik belajar dari data sebelumnya.

The purpose of this research are make application system of Artificial Neural Network and Naive Bayes to predict stroke  and to compare the accuration between of  both methods. An application applying MATLAB software has been invented to detect a stroke suspect. A good method in learning machine based on the training data is Artificial Neural Network and Naive Bayes method, by using the data variable of some common stroke symptoms used as the training data in the learning process of the system which is going to be built to determine whether prediction of stroke disease. From 120 experiments data which had been done, will be counted the accuracy of the system which divided into some training data and the other experiment data. Retrieved the percentage of  accuracy system, The Artificial Neural Network is 71,11 percent whereas Naive Bayes is 80,555 percent. Naive Bayes is more accurate than Artificial Neural Network in terms of new data decision making, but Artificial Neural Networks has better technique than Naive Bayes. Artifical Neural Network has the characteristics of learning from previous data.

Keywords

Artificial Neural Network, MATLAB, Naive Bayes, Stroke

Full Text:

PDF

References

Brunser, AM et al. 2013. Accuracy of Diffusion- Weighted Imaging in the Deteksi of Stroke in Patiens With Suspected Cerebral Infarct. Stroke American Heart Association, 44(4) : 1169-1171. Tersedia di http://stroke.ahajournals.org/content/44/4/1169 [diakses 26 April 2014].

Shanthi et al. 2010. Designing an Artificial Neural Network Model for the Prediction of Thrombo-embolic Stroke. International Journals of Biometric and Bioinformatics (IJBB). 3(1) : 1-18.

Hermawan, A. 2006. Jaringan Syaraf Tiruan Teori dan Aplikasi. Yogyakarta: ANDI.

Iqbal, M. 2009. Dasar Pengolahan Citra Menggunakan MATLAB. Department Ilmu dan Tehnologi Kelautan IPB.

Lingga, L. 2013. All About Stroke. Jakarta: Elex Media Komputindo.

Maes S et al. 2007. Credit Card. Fraud Detection Using Bayesian and Neural Networks. Prosiding Department of Computer Science Computional Modeling Lab (COMO). Brussel : Vrije Universiteit Brussel.

Subanar. 2008. Inferensi Bayesian. Jakarta: Universitas Terbuka.

Refbacks

  • There are currently no refbacks.




Creative Commons License
This work is licensed under a Creative Commons Attribution 4.0 International License.