Peramalan Harga Saham PT Telkom Indonesia Menggunakan Model Patch Time Series Transformer (PatchTST)

Sulistiyo, Yanuar Audrey (2026) Peramalan Harga Saham PT Telkom Indonesia Menggunakan Model Patch Time Series Transformer (PatchTST). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5026221074-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Pasar saham merupakan salah satu pilar dalam perekonomian modern yang digunakan oleh para ekonom dan ahli keuangan sebagai indikator kesehatan ekonomi suatu negara. Bagi investor, memprediksi harga saham secara akurat adalah kunci dalam merumuskan strategi investasi yang efektif dan meminimalisir resiko kerugian finansial. Oleh karena itu, tugas akhir ini menawarkan solusi untuk meningkatkan akurasi prediksi harga saham menggunakan model berbasis Deep Learning (DL). Model yang digunakan dalam tugas akhir ini adalah Patch Time Series Transformer (PatchTST) dengan data harga saham PT Telkom Indonesia Tbk. Metodologi tugas akhir ini menggunakan data harian dari 4 November 2009 hingga 4 November 2024 dengan fitur Open, High, Low, Close, Volume. Data harga saham diolah dengan dilakukan pembersihan, normalisasi, pembagian data, dan pembentukan sliding window. Untuk mendapat hasil optimal dilakukan penyetelan hiperparameter menggunakan algoritma Random Search, Tree-structured Parzen Estimator (TPE), dan Genetic Algorithm (GA). Untuk mengevaluasi kinerja model digunakan Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE) dan Mean Absolute Percentage Error (MAPE). Dari penelitian yang dilakukan didapat hasil model PatchTST mendapat nilai MAPE berkisar 2,31% hingga 4,48% yang mana lebih buruk jika dibanding informer tetapi lebih baik jika dibanding autoformer. Meskipun begitu PatchTST memiliki waktu komputasi tercepat yaitu berkisar antara 10 hingga 15 detik. Hasil dari tugas akhir ini diharapkan dapat menjadi alat pendukung pengambil keputusan bagi investor dalam mengatur strategi investasi.
===============================================================================================================================
Stock market is one of the pillars of the modern economy, used by economists and financial experts as an indicator of a country's economic health. For investors, accurately predicting stock prices is key to formulating effective investment strategies and minimizing the risk of financial loss. Therefore, this final project offers a solution to improve the accuracy of stock price predictions using a Deep Learning (DL)-based model. The model used in this final project is Patch Time Series Transformer (PatchTST) with PT Telkom Indonesia Tbk stock price data. The methodology of this final project uses daily data from November 4, 2009, to November 4, 2024, with Open, High, Low, Close, and Volume features. Stock price data is processed by cleaning, normalizing, dividing the data, and forming a sliding window. To obtain optimal results, hyperparameter tuning was performed using Random Search, Tree-structured Parzen Estimator (TPE), and Genetic Algorithm (GA). To evaluate the model's performance, Mean Absolute Error (MAE), Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Mean Absolute Percentage Error (MAPE) were used. The results of this study show that the PatchTST model achieved MAPE values ranging from 2.31% to 4.48%, which is worse than the Informer model but better than the Autoformer model. Nevertheless, PatchTST has the fastest computation time, ranging from 10 to 15 seconds. The results of this thesis are expected to serve as a decision-support tool for investors in formulating investment strategies.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Harga Saham, Peramalan, PatchTST, Deep Learning, Stock Price, Forecasting, PatchTST, Deep Learning
Subjects:	T Technology > T Technology (General) > T174 Technological forecasting T Technology > T Technology (General) > T57.5 Data Processing T Technology > T Technology (General) > T57.84 Heuristic algorithms.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Information System > 57201-(S1) Undergraduate Thesis
Depositing User:	Yanuar Audrey Sulistiyo
Date Deposited:	29 Jul 2026 07:59
Last Modified:	29 Jul 2026 07:59
URI:	http://repository.its.ac.id/id/eprint/138904

Actions (login required)

View Item