Pasulleri, Muhammad Ramadhani (2026) Prediksi harga saham syariah PT Aneka Tambang Tbk (ANTM) menggunakan metode Gated Recurrent Unit. Undergraduate thesis, Universitas Negeri Maulana Malik Ibrahim.
|
Text (Fulltext)
220605110147.pdf - Accepted Version Available under License Creative Commons Attribution Non-commercial No Derivatives. (4MB) |
Abstract
INDONESIA:
Fluktuasi harga saham yang dinamis menyebabkan tingginya ketidakpastian dalam aktivitas investasi di pasar modal sehingga diperlukan metode prediksi yang mampu menangkap pola pergerakan harga secara akurat. Penelitian ini bertujuan memprediksi harga saham PT Aneka Tambang Tbk (ANTM) menggunakan metode Gated Recurrent Unit (GRU), salah satu algoritma deep learning yang efektif untuk memodelkan data deret waktu (time series). Dataset yang digunakan berupa data historis harga penutupan harian dari Yahoo Finance pada periode 4 Januari 2016 hingga 31 Desember 2025 sebanyak 2.466 data. Penelitian ini menguji beberapa skenario dengan memvariasikan parameter model yaitu window size, jumlah unit GRU, batch size, serta rasio pembagian data latih dan data uji sebesar 80:20 dan 70:30. Kinerja model dievaluasi menggunakan Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), dan Directional Accuracy (DA). Hasil penelitian menunjukkan bahwa untuk prediksi nilai numerik harga saham, konfigurasi terbaik diperoleh pada Model F dengan rasio data 70:30 yang menghasilkan RMSE sebesar 71,66, MAPE 2,32%, MAE 45,93, dan DA sebesar 48,6968%. Sementara itu, untuk prediksi arah pergerakan harga saham, Model A dengan rasio data 70:30 memberikan hasil terbaik dengan nilai Directional Accuracy sebesar 49,8638%. Hasil ini menunjukkan bahwa metode GRU mampu memodelkan pola pergerakan harga saham dengan cukup baik dari sisi nilai maupun arah pergerakan harga.
ENGLISH:
Dynamic stock price fluctuations cause high uncertainty in investment activities in the capital market, requiring prediction methods that can accurately capture price movement patterns. This study aims to predict the stock price of PT Aneka Tambang Tbk (ANTM) using the Gated Recurrent Unit (GRU) method, one of the effective deep learning algorithms for modeling time series data. The dataset used consists of historical daily closing price data from Yahoo Finance for the period from January 4, 2016, to December 31, 2025, totaling 2,466 data points. This study tested several scenarios by varying the model parameters, namely window size, number of GRU units, batch size, and the ratio of training data to test data of 80:20 and 70:30. Model performance was evaluated using Root Mean Square Error (RMSE), Mean Absolute Percentage Error (MAPE), Mean Absolute Error (MAE), and Directional Accuracy (DA). The results show that for predicting numerical stock prices, the best configuration was obtained in Model F with a data ratio of 70:30, which produced an RMSE of 71.66, MAPE of 2.32%, MAE of 45.93, and DA of 48.6968%. Meanwhile, for predicting the direction of stock price movements, Model A with a data ratio of 70:30 provided the best results with a Directional Accuracy value of 49.8638%. These results indicate that the GRU method is capable of modeling stock price movement patterns quite well in terms of both value and direction of price movements.
ARABIC:
يف رامثتسلاا ةطشنأ يف نيقيلا مدع نم ةيلاع ةجرد يف مهسلأا راعسأ يف ةيكيمانيدلا تابلقتلا ببستت
ىلإ ةساردلا هذه فدهت .ةقدب راعسلأا ةكرح طامنأ طاقتلا ىلع ةرداق ؤبنت ةقيرط بلطتي امم ،لاملا سأر قوس
مهس رعسب ؤبنتلا PT Aneka Tambang Tbk (ANTM) ةقيرط مادختساب Gated Recurrent Unit
(GRU)، تانايبلا ةعومجم نوكتت .ةينمزلا لسالسلا تانايب ةجذمنل ةلاعفلا قيمعلا ملعتلا تايمزراوخ ىدحإ يهو
نم ةيخيراتلا ةيمويلا قلاغلإا راعسأ تانايب نم ةمدختسملا Yahoo Finance ىلإ 2016 رياني 4 نم ةرتفلل
رييغت للاخ نم تاهويرانيس ةدع ةساردلا هذه تربتخا .تانايب ةطقن 2466 غلبت يتلاو ، 2025 ربمسيد 31
تادحو ددعو ،ةذفانلا مجح يهو ،جذومنلا تاملعم GRU، تانايب ىإل بيردتلا تانايب ةبسنو ،ةعفدلا مجحو
يرذجلا عبرملا أطخلا طسوتم مادختساب جذومنلا ءادأ مييقت مت . 70:30 و 80:20 رابتخلاا (RMSE) طسوتمو
قلطملا يوئملا أطخلا (MAPE) قلطملا أطخلا طسوتمو (MAE) هاجتلاا ةقدو (DA). هنأ جئاتنلا ترهظأ
جذومنلا يف نيوكت لضفأ ىلع لوصحلا مت ،ةيمقرلا مهسلأا راعسأ عقوتل ةبسنلاب F امم ، 70:30 تانايب ةبسنب
ىلإ ىدأ RMSE و ، 71.66 هردق MAPE و ،٪ 2.32 هردق MAE و ، 45.93 هردق DA 48.6968٪ هردق .
جذومنلا مدق ،مهسلأا راعسأ تاكرحت هاجتا عقوتل ةبسنلاب ،هسفن تقولا يفو A لضفأ 70:30 تانايب ةبسنب
ةقيرط نأ ىإل جئاتنلا هذه ريشت .٪ 49.8638 تغلب ةيهاجتا ةقد ةميقب جئاتنلا GRU طامنأ ةجذمن ىلع ةرداق
.راعسلأا تاكرحت هاجتاو ةميق ثيح نم ديج لكشب مهسلأا راعسأ ةكرح
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Supervisor: | Nurhayati, Hani and Sari, Nur Fitriyah Ayu Tunjung |
| Keywords: | Prediksi Harga Saham; ANTM; GATED RECURRENT UNIT (GRU); Time series; Deep learning; Stock Price Prediction; ANTM; GATED RECURRENT UNIT (GRU); Time series; Deep learning; مهسلأا راعسأ عقوت ; ANTM; ةقلغملا راركتلا ةدحو (GRU); ملعتلا ،ةينمزلا لسلاسلا |
| Subjects: | 08 INFORMATION AND COMPUTING SCIENCES > 0801 Artificial Intelligence and Image Processing > 080108 Neural, Evolutionary and Fuzzy Computation |
| Departement: | Fakultas Sains dan Teknologi > Jurusan Teknik Informatika |
| Depositing User: | Muhammad Ramadhani Pasulleri |
| Date Deposited: | 18 Jun 2026 11:47 |
| Last Modified: | 18 Jun 2026 11:47 |
| URI: | http://etheses.uin-malang.ac.id/id/eprint/85613 |
Downloads
Downloads per month over past year
Actions (login required)
![]() |
View Item |
