PREDIKSI HARGA MATA UANG BINANCE COIN DENGAN METODE LONG SHORT TERM MEMORY

SIMBOLON, P.H. CAN ENJOY (2022) PREDIKSI HARGA MATA UANG BINANCE COIN DENGAN METODE LONG SHORT TERM MEMORY. Other thesis, UPN 'Veteran" Yogyakarta.

	Text ABSTRAK.pdf Download (454kB)
	Text DAFTAR ISI.pdf Download (14kB)
	Text DAFTAR PUSTAKA.pdf Download (288kB)
	Text Halaman Pengesahan Pembimbing.pdf Download (2MB)
	Text SKIRPSI_P.H. Can Enjoy Simbolon_123170065_Informatika.pdf Restricted to Repository staff only Download (1MB)
	Text COVER.pdf Download (101kB)

Abstract

vii
ABSTRAK
Binance Coin (BNB) adalah aset kripto yang dibuat oleh Binance, yang merupakan
bursa mata uang kripto terbesar di dunia. BNB dalam ekositem binance digunakan sebagai
token ulititas untuk membayar biaya perdagangan di platform Binance. BNB telah menjadi
aset kripto yang banyak digunakan sebagai aset investasi seiring dengan pesatnya
perkembangan Binance sebagai bursa pertukaran aset kripto. Binance Coin and Exchange
dibuat oleh Changpeng Zhao dan diluncurkan pada Juli 2017. Sejak itu, munculnya
Binance Coin dan dukungan pertukaran Cryptocurrency terbesar saat ini adalah Binance
Exchange, dan Binance adalah koin bunga pertama dan terbesar dalam hal omset harian.
volume perdagangan, menjadikannya salah satu sarana investasi pilihan bagi pengusaha
dan investor. Hal ini disebabkan oleh peningkatan nilai Binance Coin yang signifikan,
yang membangkitkan minat investor dan pebisnis untuk berinvestasi dan memperoleh
keuntungan. Namun, nilai Binance Coin tidak selalu naik setiap saat, ada kalanya harga
Binance Coin turun dan naik begitu tajam sehingga bisa menimbulkan kerugian. Oleh
karena itu, peramalan diperlukan untuk memprediksi harga Binance Coin.
Tahap pertama dari penelitian ini adalah mengumpulkan data dari sumber yang
dapat dipercaya dan menganalisis data yang terkumpul. Kemudian dilanjutkan ke
preprocessing data. Pada tahap preprocessing, data dibersihkan dan dinormalisasi dengan
Min-Max. Setelah data dibersihkan, maka dilakukan proses normalisasi untuk membawa
data ke skala 0 sampai 1. Setelah langkah pra-pemrosesan selesai, akan dilakukan pelatihan
metode metode Long Short Term Memory (LSTM)
Dalam penelitian ini, pelatihan LSTM menggunakan data historis Binance Coin
untuk memprediksi harga Binance Coin. Penelitian dilakukan dengan mengukur
keakuratan hasil prediksi dan efektivitas penelitian. Hasil penelitian ini menunjukkan
bahwa untuk total 500 epoch dan 50 neuron model LSTM menghasilkan akurasi prediksi
terbaik dengan akurasi Mean Absolute Percentage Error (MAPE) 0.0014 dan Mean
Squared Error (MSE) 0.277.Hasil penelitian didapatkan hasil terbaik dari model LSTM,
menggunakan kombinasi neuron terbaik dan periode terbaik dengan nilai MAPE 0.0014
dan MSE 0,277 selama 30 hari.
Kata kunci : binance coin,bnb,prediksi,long short term memory, investasi,recurrent neural
network, forecast.
viii
ABSTRACT
Binance Coin (BNB) is a crypto asset created by Binance, which is the largest
cryptocurrency exchange in the world. BNB in the binance ecosystem is used as a utility token
to pay trading fees on the Binance platform. BNB has become a crypto asset that is widely used
as an investment asset along with the rapid development of Binance as a cryptocurrency
exchange. Binance Coin and Exchange was created by Changpeng Zhao and launched in July
2017. Since then, the emergence of Binance Coin and the support of the largest Cryptocurrency
exchange today is Binance Exchange, and Binance is the first and largest interest coin in terms
of daily turnover. trading volume, making it one of the investment vehicles of choice for
entrepreneurs and investors. This is due to the significant increase in the value of Binance Coin,
which arouses the interest of investors and businesses to invest and earn profits. However, the
value of Binance Coin does not always go up all the time, there are times when the price of
Binance Coin drops and rises so sharply that it can cause losses. Therefore, forecasting is needed
to predict the price of Binance Coin.
The first stage of this research is to collect data from reliable sources and analyze the
collected data. Then proceed to data preprocessing. In the preprocessing stage, the data is
cleaned and normalized with Min-Max. After the data is cleaned, then the normalization process
is carried out to bring the data to a scale of 0 to 1. After the pre-processing step is complete,
training on the Long Short Term Memory (LSTM) method will be carried out.
In this study, the LSTM training used Binance Coin historical data to predict the price
of Binance Coin. The research was conducted by measuring the accuracy of the prediction
results and the effectiveness of the research. The results of this study indicate that for a total of
500 epochs and 50 neurons the LSTM model produces the best prediction accuracy with an
accuracy of Mean Absolute Percentage Error (MAPE) 0.00114 and Mean Squared Error (MSE)
0.277. The results obtained the best results from the LSTM model, using the best combination
of neurons and the best period with MAPE values of 0.00114 and MSE of 0.277 for 30 days.
Keywords: binance coin, bnb, prediction, long short term memory, investment, recurrent neural
network, forecast.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	binance coin, bnb, prediction, long short term memory, investment, recurrent neural network, forecast.
Subjects:	H Social Sciences > HD Industries. Land use. Labor
Divisions:	Faculty of Engineering, Science and Mathematics > School of Engineering Sciences
Depositing User:	Eko Yuli
Date Deposited:	25 Oct 2022 01:53
Last Modified:	25 Oct 2022 01:53
URI:	http://eprints.upnyk.ac.id/id/eprint/31468

Actions (login required)

View Item