Peramalan IHSG Berdasarkan Faktor Indeks Global Menggunakan Metode Long Short-Term Memory dan Gated Recurrent Unit

Lammoreno, Sultan Tanri (2023) Peramalan IHSG Berdasarkan Faktor Indeks Global Menggunakan Metode Long Short-Term Memory dan Gated Recurrent Unit. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 06311940000046-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 October 2025. Download (2MB) | Request a copy

Abstract

Investasi merupakan kegiatan menanam modal di masa sekarang dengan mengharapkan mendapatkan pengembalian di masa mendatang. Investasi dapat dilakukan dalam berbagai bentuk, salah satunya adalah saham. Investasi saham kini makin digemari oleh masyarakat Indonesia, terbukti dari meningkatnya jumlah investor dari tahun ke tahun. Saham merupakan surat kepemilikan dari suatu perusahaan yang diterbitkan oleh perusahaan untuk mendapatkan pendanaan. Saham dapat diukur performanya baik dalam satuan maupun dalam kelompok. Pengukuran performa suatu kelompok saham dapat disebut sebagai indeks. Indeks Harga Saham Gabungan merupakan indeks utama yang berada di Bursa Efek Indonesia karena
mencakup dari seluruh saham yang ada di Bursa Efek Indonesia. Long Short-Term Memory dan Gated Recurrent Unit merupakan model hasil modifikasi model Recurrent Neural Network dengan mengubah arsitektur selnya. Pembeda di antara Long Short-Term Memory dan Gated Recurrent Unit adalah Gated Recurrent Unit memiliki jumlah gate yang lebih sedikit dibandingkan Long Short-Term Memory, namun dapat memberikan model yang cukup efektif dan dapat dibandingkan dengan Long Short-Term Memory. Mean Absolute Percentage Error merupakan salah satu metode untuk mengevaluasi model terbaik. Hasil menunjukkan bahwa metode Gated Recurrent Unit memberikan nilai Mean Absolute Percentage Error yang lebih baik yaitu 0.525%, 0.5253%, 0.5255%, dan 0.5256%. Sedangkan untuk metode Long Short�Term Memory memberikan nilai Mean Absolute Percentage Error sebesar 0.5403%, 0.5445%, 0.5497%, dan 0.5577%.
============================================================
Investment is an activity of investing capital in the present with the expectation of getting returns in the future. Investments can be made in various forms, one of which is stocks. Stock investment is now increasingly favored by the Indonesian people, as evidenced by the increasing number of investors from year to year. Shares are ownership letters of a company issued by the company to obtain funding. Stocks can be measured both individually and in groups. Measuring the performance of a group of stocks can be referred to as an index. The Jakarta Composite Index is the main index on the Indonesia Stock Exchange because it includes all stocks on the Indonesia Stock Exchange. Long Short-Term Memory and Gated Recurrent Unit are modified models of the Recurrent Neural Network model by changing the cell architecture. The difference between Long Short-Term Memory and Gated Recurrent Unit is that Gated Recurrent Unit has fewer gates than Long Short-Term Memory, but can provide a model that is quite effective and comparable to Long Short-Term Memory. Mean Absolute Percentage Error is one of the methods to evaluate the best model.The results show that the Gated Recurrent Unit method provides better Mean Absolute Percentage Error values of 0.525%, 0.5253%, 0.5255%, and 0.5256%.Meanwhile, the Long Short-Term Memory method provides Mean Absolute Percentage Error values of 0.5403%, 0.5445%, 0.5497%, and 0.5577%

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Gated Recurrent Unit, Jakarta Composite Index, Long Short-Term Memory, Mean Absolute Percentage Error, Gated Recurrent Unit, IHSG, Long Short-Term Memory, Mean Absolute Percentage Error
Subjects:	Q Science > Q Science (General) > Q325.5 Machine learning. Support vector machines. Q Science > QA Mathematics > QA276 Mathematical statistics. Time-series analysis. Failure time data analysis. Survival analysis (Biometry) Q Science > QA Mathematics > QA336 Artificial Intelligence
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Actuaria > 94203-(S1) Undergraduate Thesis
Depositing User:	Sultan Tanri Lammoreno
Date Deposited:	10 Jan 2024 06:15
Last Modified:	10 Jan 2024 06:15
URI:	http://repository.its.ac.id/id/eprint/100839

Actions (login required)

View Item