https://aboutmusicschools.com https://slotmgc.com https://300thcombatengineersinwwii.com https://mobilephonesource.co.uk https://discord-servers.io https://esmark.net https://slotmgc.com https://nikeshoesinc.us https://ellisislandimmigrants.org https://holidaysanthology.com https://southaventownecenter.net https://jimgodfreydesign.com https://mckinneypaintingpros.com https://enchantedmansion.org https://mckinneypaintingpros.com https://laurabrodieauthor.com https://holidaysanthology.com https://ardictionary.com https://113.30.151.116 https://103.252.118.20 https://206.189.83.174 https://157.230.39.109 https://128.199.85.208 https://172.104.51.149 https://174.138.21.250 https://157.245.50.183 https://152.42.239.189 https://188.166.210.125 https://152.42.178.155 https://192.53.172.202 https://172.104.188.91 https://103.252.118.157 https://63.250.61.107 https://165.22.104.74

Penerapan Artificial Neural Network dengan Algoritma Backpropagation untuk Memprediksi Harga Saham

Authors

  • Widhi Aryanti Statistika, Universitas Islam Bandung

DOI:

https://doi.org/10.29313/jrs.v3i2.2953

Keywords:

Artificial Neural Network (ANN), Backpropagation, Sigmoid

Abstract

Abstract. Data can be utilized in making a decision in the future. In determining a crucial decision, an appropriate method is needed in predicting data. One of the most widely used methods in predicting data is multiple linear regression with the ordinary least squares (OLS) method, but in its implementation the data often does not meet the assumptions so that Atrificial Neural Network (ANN) is used as an alternative for predicting data because of its ability to carry out learning algorithms from data where in the process no assumptions are needed. ANN can study the patterns contained in the data through a series of interconnected layers, where each layer consists of a number of neurons that have certain weights and biases. The sigmoid activation function is used to activate or deactivate a neuron and to facilitate calculations when performing the Backpropagation Neural Network algorithm. The Backpropagation Neural Network (BNN) algorithm starts from initializing the initial weights and biases, then predicts the testing data, calculates errors, and updates the weights and biases using a gradient algorithm. From the test results, using a multilayer network, the architectural components 7-3-1 and the factor assessment are obtained with details of the number of iterations of 100, the learning rate value is 0.9 and the error tolerance is 0.01 so that the average result value of RMSE Atrificial Neural Network (ANN) is 0.01392 for the training process, the average RMSE Atrial Neural Network (ANN) is 0.01128 for the testing process and the accuracy reaches 96%.

Abstrak. Data dapat dimanfaatkan dalam membuat sebuah keputusan di masa yang akan datang. Dalam menentukan sebuah keputusan yang krusial diperlukan sebuah metode yang tepat dalam memprediksi data. Salah satu metode yang paling banyak digunakan dalam memprediksi data adalah regresi linear berganda dengan metode ordinary least square (OLS), namun dalam penerepannya sering kali data tidak memenuhi asusmi sehingga Atrificial neural network (ANN) dijadikan sebagai alternatif untuk memprediksi data karena kemampuannya dalam melakukan algoritma pembelajaran dari data dimana dalam prosesnya tidak diperlukan asumsi apapun. ANN dapat mempelajari pola-pola yang terdapat pada data melalui serangkaian layer yang saling terhubung, dimana setiap layer terdiri dari sejumlah neuron yang memiliki bobot dan bias tertentu. Fungsi aktivasi sigmoid digunakan untuk mengaktifkan atau menonaktifkan sebuah neuron serta untuk memudahkan perhitungan pada saat melakukan algoritma Backpropagation Neural Network. Algoritma Backpropagation Neural Network (BNN) dimulai dari inisialisasi bobot dan bias awal, kemudian melakukan prediksi terhadap data testing, menghitung error, dan memperbarui bobot dan bias menggunakan algoritma gradien. Dari hasil pengujian, menggunakan multilayer network diperoleh  komponen arsitektur 7-3-1 dan penilaian faktor dengan rincian jumlah iterasi 100, nilai learning rate sebesar 0,9 dan toleransi error sebesar 0,01 sehingga diperoleh nilai hasil rata-rata RMSE Atrificial neural network (ANN) sebesar 0,01392 untuk proses training, rata-rata RMSE Atrificial neural network (ANN) sebesar 0,01128 untuk proses testing dan akurasi mencapai 96%.

References

J. S. Sebayang and B. Yuniarto, “Perbandingan Model Estimasi Artificial Neural Network Optimasi Genetic Algorithm dan Regresi Linier Berganda,” Media Statistika, vol. 10, no. 1, p. 13, 2017, doi: 10.14710/medstat.10.1.13-23.

M. N. Zain, “Algoritma Artificial Neural Network dalam Klasifikasi Chest X-Rays Pasien COVID-19,” Jurnal Riset Statistika, pp. 137–144, Dec. 2022, doi: 10.29313/jrs.v2i2.1426.

Azizah, “Bunga Deposito Dan Volume Perdagangan Saham. In Jurnal keuangan dan Perbankan,” vol. 9, 2019.

X. Wang, “Application of artificial neural network to predict short-term capital flow,” ICRCCS 2009 - 2009 International Conference on Research Challenges in Computer Science, no. Lm, pp. 124–127, 2009, doi: 10.1109/ICRCCS.2009.39.

F. A. Y. Putri, “Studi Komparasi Peramalan Harga Minyak Mentah Menggunakan Metode Generalized Regression Neural Network dan Feed Forward Neural Network (Studi Kasus: Harga Minyak Mentah West Texas Intermediate),” Tugas Akhir, 2018.

Z. S. C. Viqri and E. Kurniati, “Perbandingan Penerapan Metode Fuzzy Time Series Model Chen-Hsu dan Model Lee dalam Memprediksi Kurs Rupiah terhadap Dolar Amerika,” DataMath: Journal of Statistics and Mathematics, vol. 1, no. 1, pp. 19–26, 2023, doi: 10.29313/datamath.v1i1.12.

N. A. K. Rifai, “Pendekatan Regresi Nonparametrik dengan Fungsi Kernel untuk Indeks Harga Saham Gabungan,” Statistika, vol. 19, no. 1, pp. 53–61, 2019.

T. Hendrawati, “Kajian Metode Imputasi dalam Menangani Missing Data,” Prosiding Seminar Nasional Matematika dan Pendidikan Matematika UMS, pp. 637–642, 2015.

C. Saranya and G. Manikandan, “A study on normalization techniques for privacy preserving data mining,” International Journal of Engineering and Technology, vol. 5, no. 3, pp. 2701–2704, 2013.

D. N. Agus Perdana Windarto, M. S. H. Anjar Wanto, Frinto Tambunan, M. R. L. Muhammad Noor Hasan Siregar, and D. N. Solikhun, Yusra Fadhillah, Jaringan Saraf Tiruan: Algoritma Prediksi dan Implementasi, vol. 53, no. 9. 2019.

Julpan, E. B. Nababan, and M. Zarlis, “Analisis Fungsi Aktivasi Sigmoid Biner dan Sigmoid Bipolar dalam Algoritma Backpropagation pada Prediksi Kemampuan Siswa,” Jurnal Teknovasi, vol. 02, pp. 103–116, 2015.

T. Chai and R. R. Draxler, “Root mean square error (RMSE) or mean absolute error (MAE)? -Arguments against avoiding RMSE in the literature,” Geosci Model Dev, vol. 7, no. 3, pp. 1247–1250, 2014, doi: 10.5194/gmd-7-1247-2014.

Downloads

Published

2023-12-24