SUPRIYADI, SUPRIYADI (2025) PREDIKSI HARGA SAHAM MENGGUNAKAN METODE LONG SHORT-TERM MEMORY (LSTM) BERDASARKAN DATA HISTORIS PERDAGANGAN SAHAM. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
01 COVER.pdf Download (530kB) | Preview |
|
![]() |
Text (BAB I)
02 BAB 1.pdf Restricted to Registered users only Download (61kB) |
|
![]() |
Text (BAB II)
03 BAB 2.pdf Restricted to Registered users only Download (157kB) |
|
![]() |
Text (BAB III)
04 BAB 3.pdf Restricted to Registered users only Download (115kB) |
|
![]() |
Text (BAB IV)
05 BAB 4.pdf Restricted to Registered users only Download (1MB) |
|
![]() |
Text (BAB V)
06 BAB 5.pdf Restricted to Registered users only Download (34kB) |
|
![]() |
Text (DAFTAR PUSTAKA)
07 DAFTAR PUSTAKA.pdf Restricted to Registered users only Download (98kB) |
|
![]() |
Text (LAMPIRAN)
08 LAMPIRAN.pdf Restricted to Registered users only Download (634kB) |
Abstract
He volatile nature of stock price movements and the multitude of influencing factors make stock price prediction both a challenge and a necessity in investment decision-making. This study aims to examine the effectiveness of the Long Short-Term Memory (LSTM) model in predicting the stock price of PT Aneka Tambang Tbk (ANTM) based on historical trading data. The primary research question is to what extent the LSTM model can learn historical patterns (OHLCV) and generate accurate closing price predictions. This research adopts a quantitative approach with an experimental design, utilizing daily stock data of ANTM from 2020 to 2024 sourced from the Indonesia Stock Exchange. The data were processed using the Python programming language on the Google Colaboratory platform, involving MinMaxScaler normalization, sequence creation with a 50-time-step window, and a two-layer LSTM architecture. Model performance was evaluated using MAE, RMSE, MAPE, MSE, and R² metrics. The results showed that the LSTM model achieved a MAPE of 4,24% and an R² of 0,6646, indicating a good level of accuracy for shortterm prediction. These findings strengthen the relevance of deep learning in stock market analysis based on historical data. In conclusion, LSTM has proven effective in modeling the historical behavior of ANTM’s stock price and can be further enhanced by incorporating external variables and applying it to other stocks to improve prediction accuracy and generalizability. Keywords: Stock Prediction, LSTM, Historical, Deep Learning, ANTM. Pergerakan harga saham yang bersifat fluktuatif dan dipengaruhi banyak faktor menjadikan prediksi harga saham sebagai tantangan sekaligus kebutuhan penting dalam pengambilan keputusan investasi. Penelitian ini bertujuan untuk mengkaji efektivitas model Long Short-Term Memory (LSTM) dalam memprediksi harga saham PT Aneka Tambang Tbk (ANTM) berdasarkan data historis perdagangan saham. Pertanyaan utama yang ingin dijawab adalah sejauh mana model LSTM mampu mempelajari pola data historis (OHLCV) dan menghasilkan prediksi harga penutupan yang akurat. Penelitian ini menggunakan pendekatan kuantitatif dengan desain eksperimental, memanfaatkan dataset harian saham ANTM periode 2020- 2024 yang diperoleh dari Bursa Efek Indonesia. Data dianalisis menggunakan bahasa pemrograman Python melalui platform Google Colaboratory, dengan preprocessing berupa normalisasi MinMaxScaler, pembentukan sequence 50 langkah waktu, dan arsitektur LSTM dua lapis. Evaluasi performa dilakukan dengan metrik MAE, RMSE, MAPE, MSE, dan R². Hasil penelitian menunjukkan bahwa model LSTM mampu menghasilkan MAPE sebesar 4,24% dan R² sebesar 0,6646, yang mengindikasikan tingkat akurasi yang baik dalam prediksi jangka pendek. Temuan ini memperkuat relevansi penggunaan deep learning dalam analisis pasar saham berbasis data historis. Kesimpulannya, LSTM terbukti efektif untuk memodelkan hubungan historis harga saham ANTM dan dapat dikembangkan lebih lanjut dengan menambahkan variabel eksternal serta diuji pada saham lain guna meningkatkan generalisasi dan akurasi prediksi. Kata kunci: Prediksi Saham, LSTM, Data Historis, Deep Learning, ANTM.
Actions (login required)
![]() |
View Item |