PERBANDINGAN KINERJA JARINGAN SYARAF TIRUAN DAN FUZZY INFERENCE SYSTEM UNTUK PREDIKSI PRESTASI PESERTA DIDIK

SITI HELMIYAH, NIM. 12650003 (2016) PERBANDINGAN KINERJA JARINGAN SYARAF TIRUAN DAN FUZZY INFERENCE SYSTEM UNTUK PREDIKSI PRESTASI PESERTA DIDIK. Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (PERBANDINGAN KINERJA JARINGAN SYARAF TIRUAN DAN FUZZY INFERENCE SYSTEM UNTUK PREDIKSI PRESTASI PESERTA DIDIK) 12650003_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf Download (3MB) | Preview
	Text (PERBANDINGAN KINERJA JARINGAN SYARAF TIRUAN DAN FUZZY INFERENCE SYSTEM UNTUK PREDIKSI PRESTASI PESERTA DIDIK) 12650003_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (2MB)

Abstract

Achievement is a result of someone who excels in any field. In the educational world, achievement is often associated with academic value. This value is then used as a reference of a student to be called excellent academically. Manual data processing takes long time. It is necessary to use the achievements of predictive computing system that would be helpful for the prediction process. Computing system that is often used in making predictions is a neural network and fuzzy inference system. Data used in the study were taken from MAN Model Palangkaraya form of eleven subjects of UAS when MTs value and the average value of one semester report cards when the Supreme Court. The data for input to the neural network backpropagationdinormalisasikan with small intervals are [0.1, 0.9] because it uses binary sigmoid activation function, and for data to be entered into the fuzzy inference system is the original data is multiplied by 10. After the data is processed, the data do at the stage of testing using neural networks and fuzzy inference system which will compare the results obtained. Based on data obtained and which have been tested in this study, the percentage of learners' achievements prediction on back propagation neural network in the training and validation process to produce a percentage of 100% with one hidden layer architecture, the optimal parameters MSE = 0.0001, learning rate = 0 , 9, momentum = 0.4. As for the prediction of learners' achievements in the fuzzy inference system mamdani method by using S curve and bell curve (PI curve) to produce a percentage of 83.8%.

Share this knowledge with your friends :

Actions (login required)

View Item