Prediksi Harga Bitcoin dengan Pendekatan Deep Learning Menggunakan Algoritma LSTM

Muhammad, Fathur Rahmansyah Maulana (2026) Prediksi Harga Bitcoin dengan Pendekatan Deep Learning Menggunakan Algoritma LSTM. Undergraduate thesis, UPN Veteran Jawa Timur.

	Text (Cover) 21082010073.-cover.pdf Download (2MB)
	Text (Bab 1) 21082010073.-bab1.pdf Download (40kB)
	Text (Bab 2) 21082010073.-bab2.pdf Restricted to Repository staff only until 20 January 2029. Download (311kB)
	Text (Bab 3) 21082010073.-bab3.pdf Restricted to Repository staff only until 20 January 2029. Download (560kB)
	Text (Bab 4) 21082010073.-bab4.pdf Restricted to Repository staff only until 20 January 2029. Download (1MB)
	Text (Bab 5) 21082010073.-bab5.pdf Download (133kB)
	Text (Daftar Pustaka) 21082010073.-daftarpustaka.pdf Download (391kB)
	Text (Lampiran) 21082010073.-lampiran.pdf Restricted to Repository staff only until 20 January 2029. Download (347kB)

Abstract

The advancement of digital technology has catalyzed numerous innovations within the global financial sector, prominently featuring cryptocurrency with Bitcoin as the asset with the highest market capitalization. However, the extreme price volatility of Bitcoin poses a significant challenge for investors in forecasting market trends. This research aims to design a Bitcoin price prediction model utilizing the Long Short-Term Memory (LSTM) algorithm through the evaluation of six experimental scenarios. The dataset comprises historical Bitcoin prices spanning from 2014 to 2025, processed through Min-Max Scaling normalization and time-series windowing techniques. Experimental results identify Scenario 4 as the most optimal model, featuring a two layer stacked LSTM architecture with 100 hidden units each and a 120 day observation window (time step). Model evaluation yielded a Root Mean Squared Error (RMSE) of 1915.53, indicating a high level of predictive accuracy relative to Bitcoin's price scale. Visual analysis confirms the model's capability to consistently follow major price trends, although limitations remain in capturing extreme price spikes. The model is implemented via a Flask based web interface to provide interactive prediction services. This study is expected to serve as a data driven reference for investment decision-making in the digital asset market.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Permatasari, Reisa	NIDN0014059203	reisa.permatasari.sifo@upnjatim.ac.id
Thesis advisor	Najaf, Abdul Rezha Efrat	NIDN0029099403	rezha.efrat.sifo@upnjatim.ac.id

Subjects:

T Technology > T Technology (General)

Divisions:

Faculty of Computer Science > Departemen of Information Systems

Depositing User:

Fathur Rahmansyah Muhammad Maulana

Date Deposited:

22 Jan 2026 08:36

Last Modified:

22 Jan 2026 08:36

URI:

https://repository.upnjatim.ac.id/id/eprint/48929