Claim Missing Document
Check
Articles

Found 1 Documents
Search

PENERAPAN ALGORITMA BACKPROPAGATION DENGAN ADAM OPTIMIZER DALAM MEMPREDIKSI HARGA BITCOIN TERHADAP USD Adi Andrian; Oni Soesanto; Sigit Dwi Prabowo
RAGAM: Journal of Statistics & Its Application Vol 3, No 2 (2024): RAGAM: Journal of Statistics & Its Application
Publisher : Universitas Lambung Mangkurat

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.20527/ragam.v3i2.13354

Abstract

Bitcoin is a currency that implements an online transaction system without involving banks. To investors, Bitcoin is seen as a promising investment instrument due to its consistent price increase every year. However, it is important to note that Bitcoin is also a high-risk investment instrument, requiring specific techniques to consider when making buy or sell decisions. Backpropagation is a method in Artificial Neural Networks known for its good ability to generate predictions. This method can adjust network weights to reduce prediction errors and does not require assumption testing to apply this method to data. The aim of this research is to implement the Backpropagation algorithm with the Adam Optimizer to predict Bitcoin prices against USD. This method will perform computational calculations to produce predictions close to the actual values. The research results in an optimal model with 5 input layers, 5 hidden layers, and 1 output layer. The training and testing data are divided with a ratio of 70% to 30%, and the maximum number of epochs used is 3000. The accuracy results using the backpropagation method optimized with the Adam Optimizer produced predictions with an average value of 46,346.91, a MAPE of 0.6962%, a MSE of 160,159.53, and a RMSE of 400.20.