scholarly journals Pemanfaatan Data Mining dalam Penentuan Rekomendasi Mustahik (Penerima Zakat)

2020 ◽  
Vol 12 (2) ◽  
pp. 67-73
Author(s):  
Andi Abdul malik Ahmad ◽  
Zawiyah Saharuna ◽  
Muhammad Fajri Raharjo
Keyword(s):  
Neural Network ◽  
Data Mining ◽  
Artificial Neural Network ◽  
Learning Rate ◽  
Test Results ◽  
Network Algorithm ◽  
Neural Network Algorithm ◽  
Artificial Neural ◽  
Hidden Layer ◽  
Artificial Neural Network Algorithm

This study applies data mining in determining recommendations for mustahik. The application is carried out using a classification method with an artificial neural network algorithm where the attributes used are age and type of work of the head of the family, the condition and ownership of the residence, the place of sewage, family monthly income, number of dependents, and diet. Tests are carried out using a combination of values ​​between learning rate, epoch, k-fold, and hidden layer neurons. Based on the test results from the classification process, it is found that the artificial neural network algorithm has the highest accuracy when the number of hidden layer neurons is six, the learning rate is one, the fold is seven, and the number of epochs is 200, which is 92.09%. The test results are then displayed on the Mustahik information system page.

scholarly journals Diagnosis of Tuberculosis By Artificial Neural Network Algorithm

SinkrOn ◽  
2019 ◽  
Vol 3 (2) ◽  
pp. 223
Author(s):  
Amrin Amrin
Keyword(s):  
Neural Network ◽  
Data Mining ◽  
Artificial Neural Network ◽  
Cross Validation ◽  
Confusion Matrix ◽  
Network Algorithm ◽  
Neural Network Algorithm ◽  
Artificial Neural ◽  
Artificial Neural Network Algorithm

Sangat penting bagi dokter untuk melakukan diagnosa secara dini penyakit tuberculosis agar dapat mengurangi penularan penyakit tersebut kepada masyarakat luas.  Pada penelitian ini, penulis akan menerapkan metode klasifikasi data mining, yaitu Algoritma Jaringan Syaraf Tiruan untuk mendiagnosa penyakit tuberculosis. Berdasarkan hasil pengukuran performa dari model tersebut dengan  menggunakan  metode pengujian Cross Validation, Confusion Matrix dan Kurva ROC, diketahui bahwa algoritma jaringan syaraf tiruan memiliki tingkat akurasi sebesar 89,89% dan nilai area under the curva (AUC) sebesar 0,975. Hal ini menunjukkan bahwa model yang dihasilkan termasuk katagori klasifikasi  sangat baik karena memiliki nilai AUC antara 0.90-1.00.

Sign in / Sign up

Export Citation Format

Share Document