PREDIKSI HARGA SAHAM SYARIAH MENGGUNAKAN METODE LONG SHORT TERM MEMORY (LSTM)

Muhammad Fauzi Romadon, NIM.: 17106010025 (2023) PREDIKSI HARGA SAHAM SYARIAH MENGGUNAKAN METODE LONG SHORT TERM MEMORY (LSTM). Skripsi thesis, UIN SUNAN KALIJAGA YOGYAKARTA.

Preview	Text (PREDIKSI HARGA SAHAM SYARIAH MENGGUNAKAN METODE LONG SHORT TERM MEMORY (LSTM)) 17106010025_BAB-I_IV-atau-V_DAFTAR-PUSTAKA.pdf - Published Version Download (2MB) | Preview
	Text (PREDIKSI HARGA SAHAM SYARIAH MENGGUNAKAN METODE LONG SHORT TERM MEMORY (LSTM)) 17106010025_BAB-II_sampai_SEBELUM-BAB-TERAKHIR.pdf Restricted to Registered users only Download (3MB) | Request a copy

Abstract

Stocks are one of the attractive investment options because with good prediction and processing of funds investors will get big profits, but because the movement of the stock curve is very dynamic, it makes it difficult to predict stock prices. Historical stock price data is a series of stock data variables arranged in a certain event from time to time which is classified as time series data. Time series data can be analyzed using time series analysis with several methods, this study uses the Long Short Term Memory (LSTM) method. LSTM combined with analytical parameters such as the amount of data, hidden neurons, max epoch and the best optimaizer, will be analyzed with several sharia stocks for the period 1 January 2016 to 31 December 2020 including shares of Telekomunikasi Indonesia Tbk, Sarana Menara Nusantara Tbk, XL Axiata Tbk as material. optimal model formation test and PT Indosat Tbk as a test of the model created. Obtained with 3th data or 756 historical stock data, 50 hidden neurons, 50 epochs, and the ADAMax optimaizer produces accurate stock price predictions with a MAPE value of less than 10% in the formation of an optimal model and a MAPE value of 3.27% in testing model.

Share this knowledge with your friends :

Actions (login required)

View Item