Implementasi Model Long Short Term Memory (LSTM) Pada Proyeksi Harga Saham (Studi Kasus: PT. Pertamina Geothermal Energy (Persero))

Authors

  • Dian Christien Arisona Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia
  • Agusrawati Agusrawati Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia
  • Makkulau Makkulau Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia
  • Irma Yahya Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia
  • Gusti Ngurah Adhi Wibawa Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia
  • Baharuddin Baharuddin Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia
  • Putri Riski Fahyuni Program Studi Statistika, Fakultas Matematika dan Ilmu pengetahuan Alam, Universitas Halu Oleo, Kendari, 93232, Indonesia

DOI:

https://doi.org/10.20956/ejsa.v6i2.44963

Keywords:

Stock Price Forecasting, Time Series, LSTM, Financial Analysis, Machine Learning

Abstract

This research presents a comprehensive analysis of the Long Short Term Memory (LSTM) method in projecting the stock price of PT. Pertamina Geothermal Energy (Persero). Utilizing daily stock price data, the LSTM model achieves a high level of accuracy with a Mean Absolute Percentage Error (MAPE) value of 0.84%. The LSTM's gate mechanism (input, forget, output) enables it to store long-term information, controlling the flow of information to update memory, delete irrelevant data, and generate predictions. Optimized with backpropagation through time (BPTT) and activation functions, the LSTM model proves effective in investment decision making, providing valuable insights for investors and market players to anticipate stock price fluctuations. This research demonstrates the great potential of machine learning in financial analysis, particularly in stock price projection and time series analysis. The results indicate that LSTM can be a valuable tool for investors and financial analysts, enhancing their ability to make informed decisions.

References

Widiyanto, M. H., Mayasari, R., & Garno. Implementasi Time Series Pada Data Penjualan Di Gaikindo Menggunakan Algoritma Seasonal ARIMA. Jurnal Mahasiswa Teknik Informatika, 7(3), 1501-1506, 2023.

Arwansyah, A., Suryani, S., Hasrif, S. Y., Usman, U., Ahyuna, A., & Samsu Alam. Time Series Forecasting Menggunakan Deep Gated Recurrent Units. Digital Transformation Technology, 4(1), 410–416, 2022.

Yulisa, P. N., Haris, M., & Arum, P. R. Peramalan Nilai Ekspor Migas di Indonesia dengan Model Long Short Term Memory(LSTM) dan Gated Recurrent Unit(GRU). Jurnal Statistika. 16(1), 328 -341, 2023.

Karyadi, Y. Prediksi Kualitas Udara Dengan Metoda LSTM, Bidirectional LSTM, dan GRU. JATISI (Jurnal Teknik Informatika Dan Sistem Informasi), 9(1), 671-684, 2022.

Anggraeni, D. T. Forecasting Harga Saham Menggunakan Metode Simple Moving Average Dan Web Scrapping. Jurnal Ilmiah Matrik, 21(3), 234-241, 2019.

Lamabelawa, M. I. J. Analisis Perhitungan Metode Interpolasi Pada Data Time Series Kemiskinan Di NTT. Jurnal (HOAQ) – Teknologi Informasi, 8(1), 640-646, 2018.

Rahmah, A. L. Analisis Model Multivariate Long Short-Term Memory Untuk Prakiraan Kualitas Udara Dki Jakarta Berdasarkan Data Tahun 2010-2022. Jakarta: Program Studi Sistem Informasi Fakultas Sains Dan Teknologi Universitas Islam Negeri Syarif Hidayatullah, 2024.

Ramadhan, M. F. Prediksi Harga Bitcoin Menggunakan Metode Long Short Term Memory. Jambi: Program Studi Sistem Informasi Fakultas Sains Dan Teknologi Universitas Jambi, 2024.

Downloads

Published

2025-08-04

Issue

Vol. 6, No. 2, Juli, 2025 : Estimasi

Section

Articles

License

Copyright

It is the author's responsibility to ensure that his or her submitted work does not infringe any existing copyright. Authors should obtain permission to reproduce or adapt copyrighted material and provide evidence of approval upon submitting the final version of a manuscript.