Haaniyah Yarnida, Hani (2019) IMPLEMENTASI JARINGAN SARAF TIRUAN UNTUK MEMPREDIKSI NILAI MATA PELAJARAN UJIAN NASIONAL SMK NEGERI 2 CIMAHI DENGAN METODE BACKPROPAGATION. S1 thesis, Universitas Pendidikan Indonesia.
![[img]](http://repository.upi.edu/style/images/fileicons/text.png) |
Text
S_PTE_1504674_Title.pdf Download (473kB) |
|
Text
S_PTE_1504674_Chapter1.pdf Download (110kB) |
|
Text
S_PTE_1504674_Chapter2.pdf Restricted to Staf Perpustakaan Download (155kB) |
|
Text
S_PTE_1504674_Chapter3.pdf Download (536kB) |
|
Text
S_PTE_1504674_Chapter4.pdf Restricted to Staf Perpustakaan Download (473kB) |
|
Text
S_PTE_1504674_Chapter5.pdf Download (104kB) |
|
Text
S_PTE_1504674_Appendix.pdf Restricted to Staf Perpustakaan Download (24MB) |
Abstract
Penelitian ini bertujuan untuk mengetahui prediksi nilai Ujian Nasional di SMKN 2 Cimahi dengan menggunakan metode backpropagation. Penggunaan metode backpropagation pada penelitian ini agar metode terbaik untuk mendapatkan hasil yang optimal karena memiliki akurasi tinggi. Untuk mengetahui pengaruh akurasi prediksi terhadap input pembelajaran, maka pelatihan jaringan dilakukan sebanyak dua kali. Semua data diolah menggunakan metode backpropagation. Hasil prediksi 2017-2018 dengan menggunakan data tahun 2011-2016, menggunakan arsitektur jaringan dengan hidden layer 5 dan 10 dengan nilai MSE 0,001 dengan maksimal epoch 1000. Pelatihan terbaik yaitu dengan menggunakan 5 hidden layer karena menghasilkan epoch yang paling kecil dan akurasi yang besar. Pelatihan berhenti di iterasi ke-127 dengan nilai epoch 0,0024058 dengan akurasi sebesar 96,50%. Pada gambar hasil prediksi menunjukkan perbandingan antara target dengan keluaran JST pada data pengujian. Pada gambar tersebut dapat dilihat bahwa sebagian target dan keluaran jaringan sudah berdekatan (hampir menempati posisi yang sama). Prediction of test scores is important to know the level of graduation, therefore we need a system that can predict student test scores so that the school can take action to improve the learning of students that aims to increase the number of students graduating each year. Predictions about the value of the National Examination need to be known early on as a representative step of the ability of students and can be used as a benchmark for the achievement of the value of the National Examination while. Especially at the level of certain subjects that the level of difficulty is relatively higher than other subjects. This study aims to determine the prediction value of the National Examination at SMK 2 Cimahi by using the backpropagation method. The use of the backpropagation method in this study is for the best method to get optimal results because it has high accuracy. To find out the effect of prediction accuracy on learning inputs, network training is carried out twice. All data is processed using the backpropagation method. The 2017-2018 prediction results using 2011-2016 data, using network architecture with hidden layers 5 and 10 with an MSE value of 0.001 with a maximum of epoch 1000. The best training is to use 5 hidden layers because it produces the smallest epoch and great accuracy. The training stopped at 127th iteration with the epoch value of 0.0024058 with an accuracy of 96.50%. Comparison of prediction results with the original 2018 data values resulted in an accuracy of 95.28%
| Item Type: | Thesis (S1) |
|---|---|
| Uncontrolled Keywords: | Prediksi, backpropagation, Ujian Nasional |
| Subjects: | L Education > L Education (General) L Education > LB Theory and practice of education > LB1603 Secondary Education. High schools T Technology > TK Electrical engineering. Electronics Nuclear engineering |
| Divisions: | Fakultas Pendidikan Teknologi dan Kejuruan > Jurusan Pendidikan Teknik Elektro > Program Studi Pendidikan Teknik Elektro |
| Depositing User: | Haaniyah Yarnida |
| Date Deposited: | 03 Feb 2020 06:09 |
| Last Modified: | 03 Feb 2020 06:09 |
| URI: | http://repository.upi.edu/id/eprint/38826 |
Actions (login required)
 |
View Item |