Prediksi Harga Saham PT United Tractors Tbk. Menggunakan Support Vector Regression (SVR) dengan Algoritma Particle Swarm Optimization (PSO)

Abdilah, Haykal Hafiz (2025) Prediksi Harga Saham PT United Tractors Tbk. Menggunakan Support Vector Regression (SVR) dengan Algoritma Particle Swarm Optimization (PSO). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5003211126-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (2MB) | Request a copy

Abstract

Prediksi harga saham menjadi aspek krusial dalam pengambilan keputusan investasi di pasar modal. Penelitian kali ini berfokus untuk pemodelan prediksi harga saham PT United Tractors (UNTR) menggunakan metode Support Vector Regression (SVR) yang di optimasi menggunakan algoritma Particle Swarm Optimization (PSO). Penggunaan algoritma PSO digunakan dalam memperoleh nilai optimal dari parameter dan meminimalkan proses komputasi. Data yang digunakan mencakup nilai close price harian saham UNTR yang dicatat secara harian berdasarkan kalender Bursa Efek Indonesia dari tanggal 01 Januari 2022 hingga 31 Desember 2024. Pendekatan multivariat diterapkan dengan memanfaatkan nilai harga saham UNTR periode sebelumnya sebagai variabel independen. Variabel independen lainnya adalah nilai harian saham ASII dan IHSG para periode sebelumnya. Penelitian ini menghasilkan data bahwa pemodelan prediksi dengan SVR-PSO menghasilkan nilai MAPE 0,01197, lebih rendah dibandingkan dengan pemodelan prediksi dengan SVR-GridSearch yang menghasilkan MAPE 0,0124. Daya komputasi dari SVR-PSO juga lebih efisien dengan memakan waktu hanya 6 detik untuk menghasilkan pemodelannya, jauh dibanding SVR-GridSearch yang memerlukan waktu 120 detik untuk menghasilkan pemodelannya. Pemanfaatan algoritma machine learning terlihat dapat memaksimalkan hasil prediksi untuk studi kasus harga saham UNTR.
============================================================================================================================================
Stock price prediction is a crucial aspect in making investment decisions in the capital market. This research focuses on modeling the prediction of PT United Tractors (UNTR) stock prices using the Support Vector Regression (SVR) method which is optimized using the Particle Swarm Optimization (PSO) algorithm. The use of the PSO algorithm is used in obtaining the optimal value of the parameters and minimizing the computational process. The data used includes the daily close price value of UNTR shares recorded daily based on the Indonesia Stock Exchange calendar from January 01, 2022 to December 31, 2024. A multivariate approach is applied by utilizing the previous period's UNTR stock price value as an independent variable. Other independent variables are the daily value of ASII and JCI stocks in the previous period. This research produces data that prediction modeling with SVR-PSO produces a MAPE value of 0.01197, lower than prediction modeling with SVR-Grid Search which produces a MAPE of 0.0124. The computational power of SVR-PSO is also more efficient by taking only 6 seconds to produce the modeling, compared to SVR-Grid Search which takes 120 seconds to produce the modeling. The use of machine learning algorithms can maximize the prediction results for the UNTR stock price case study.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Particle Swarm Optimization, Prediksi, Saham, Support Vector Regression, United Tractors, Prediction, Stock.
Subjects:	H Social Sciences > HA Statistics > HA30.3 Time-series analysis H Social Sciences > HG Finance > HG4915 Stocks--Prices Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > Q Science (General) > Q337.3 Swarm intelligence
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Statistics > 49201-(S1) Undergraduate Thesis
Depositing User:	Haykal Hafiz Abdilah
Date Deposited:	31 Jul 2025 08:07
Last Modified:	31 Jul 2025 08:07
URI:	http://repository.its.ac.id/id/eprint/124796

Actions (login required)

View Item