THE USE OF PCA TO OVERCOME THE MULTICOLLINEARITY IN ARIMAX MODEL FOR FORECASTING THE RUPIAH-USD EXCHANGE RATE

NAJLA KHANSA, ALIFAH (2026) THE USE OF PCA TO OVERCOME THE MULTICOLLINEARITY IN ARIMAX MODEL FOR FORECASTING THE RUPIAH-USD EXCHANGE RATE. FAKULTAS MATEMATIKA DAN ILMU PENGETAHUAN ALAM, UNIVERSITAS LAMPUNG.

[img]
Preview
 Text
ABSTRAK.pdf
Download (38kB) | Preview
[img] Text
SKRIPSI FULL.pdf
Restricted to Repository staff only
Download (5MB) | Request a copy
[img]
Preview
 Text
SKRIPSI TANPA BAB PEMBAHASAN.pdf
Download (4MB) | Preview

Abstract

Peramalan deret waktu digunakan untuk memperkirakan nilai suatu variabel di masa mendatang berdasarkan pola historisnya. Dalam praktiknya, variabel tidak hanya dipengaruhi oleh nilai masa lalu, tetapi juga oleh faktor eksternal, sehingga digunakan model Autoregressive Integrated Moving Average with Exogenous Variables (ARIMAX). Namun, penggunaan beberapa variabel eksogen berpotensi menimbulkan multikolinearitas yang dapat menurunkan keandalan model. Penelitian ini bertujuan menerapkan Principal Component Analysis (PCA) untuk mengatasi multikolinearitas pada model ARIMAX dalam meramalkan nilai tukar rupiah terhadap dolar Amerika Serikat periode Januari hingga Desember 2025. Data yang digunakan merupakan data sekunder bulanan periode Januari 2012 hingga Desember 2024, dengan variabel eksogen berupa jumlah uang beredar, suku bunga (BI Rate), inflasi, ekspor, dan impor. Hasil penelitian menunjukkan bahwa PCA menghasilkan dua komponen utama yang mampu menjelaskan 85.55% keragaman data. Model terbaik adalah ARIMAX(0,1,2) dengan transformasi logaritma natural. Hasil peramalan menunjukkan bahwa nilai tukar rupiah berfluktuasi namun cenderung mengalami depresiasi, dengan tingkat kesalahan pada setiap periode berkisar antara 0.15546% hingga 3.716329% dan nilai Mean Absolute Percentage Error (MAPE) sebesar 1.263363%, sehingga model memiliki tingkat akurasi yang tinggi. Keywords: ARIMAX, Multikolinearitas, Nilai Tukar Rupiah, Peramalan, Principal Component Analysis. Time series forecasting is used to estimate the future value of a variable based on its historical patterns. In practice, a variable is not only influenced by its past values but also by external factors, leading to the use of the Autoregressive Integrated Moving Average with Exogenous Variables (ARIMAX) model. However, the inclusion of multiple exogenous variables may lead to multicollinearity, which can reduce the reliability of the model. This study aims to apply Principal Component Analysis (PCA) to address multicollinearity in the ARIMAX model for forecasting the Rupiah exchange rate against the United States Dollar for the period January to December 2025. The data used are monthly secondary data from January 2012 to December 2024, with exogenous variables including money supply, interest rate, inflation, exports, and imports. The results show that PCA produces two principal components that explain 85.55% of the total data variability. The best model obtained is ARIMAX(0,1,2) with natural logarithm transformation. The forecasting results indicate that the Rupiah exchange rate fluctuates but tends to depreciate, with forecasting errors in each period ranging from 0.15546% to 3.716329% and a Mean Absolute Percentage Error (MAPE) of 1.263363%, indicating that the model has a high level of accuracy. Keywords: ARIMAX, Rupiah Exchange Rate, Forecasting, Multicollinearity, Principal Component Analysis.

Item Type: Other
Subjects: ?? 500 ??
?? 510 ??
Divisions: Fakultas MIPA > Prodi Matematika
Depositing User: 2605720058 Digilib
Date Deposited: 24 Jun 2026 03:30
Last Modified: 24 Jun 2026 03:30
URI: http://digilib.unila.ac.id/id/eprint/101373

Actions (login required)

View Item View Item