Tejakusuma, Christina Erica Sugianto (2022) Analisis kurs mata uang asing terhadap nilai ekspor-impor Indonesia dengan metode partial least square regression = Foreign exchange rate towards Indonesian export-import value analysis using partial least square regression method. Bachelor thesis, Universitas Pelita Harapan.
|
Text (Title)
Title.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (211kB) | Preview |
|
|
Text (Abstract)
Abstract.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (590kB) | Preview |
|
|
Text (ToC)
ToC.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (556kB) | Preview |
|
![]() |
Text (Chapter1)
Chapter1.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (547kB) |
|
![]() |
Text (Chapter2)
Chapter2.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (718kB) |
|
![]() |
Text (Chapter3)
Chapter3.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (880kB) |
|
![]() |
Text (Chapter4)
Chapter4.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (1MB) |
|
![]() |
Text (Chapter5)
Chapter5.pdf Restricted to Registered users only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (554kB) |
|
|
Text (Bibliography)
Bibliography.pdf Available under License Creative Commons Attribution Non-commercial Share Alike. Download (530kB) | Preview |
|
![]() |
Text (Appendices)
Appendices.pdf Restricted to Repository staff only Available under License Creative Commons Attribution Non-commercial Share Alike. Download (2MB) |
Abstract
Pada masalah multikolinearitas, variabel bebas yang berkorelasi tinggi dengan variabel bebas lainnya umumnya dihilangkan, padahal variabel tersebut cukup memiliki pengaruh terhadap variabel tak bebas. Untuk itu, dilakukan penelitian agar masalah multikolinearitas dapat diatasi tanpa perlu menghilangkan variabel-variabelnya. Regresi Partial Least Square (PLS) dapat menjadi solusi masalah ini dan akan diaplikasikan pada analisis kurs mata uang asing terhadap nilai ekspor-impor Indonesia yang memiliki masalah multikolinearitas. Data yang digunakan merupakan data selama 120 bulan dari Januari 2012 hingga Desember 2021. Data ini akan dibagi ke dalam data train dan data test dengan perbandingan 90:10, kemudian data ditransformasi menggunakan metode Cochrane-Orcutt. Regresi PLS diawali dengan uji signifikansi untuk pembentukan komponen PLS, di mana pembentukan komponen yang baru akan berhenti ketika sudah tidak ada lagi variabel yang signifikan. Kemudian komponen baru yang terbentuk akan diregresi terhadap variabel tak bebas untuk memperoleh model PLS. Hasil model PLS pada penelitian ini berhasil menghilangkan multikolinearitas dan memenuhi seluruh asumsi klasik. Hasil yang diperoleh tersebut juga dibandingkan dengan hasil pada model Ordinary Least Square (OLS) dengan metode backward elimination. Nilai Root Mean Square Error (RMSE) yang diperoleh pada model PLS lebih kecil dibandingkan pada model OLS, sedangkan nilai adjusted R² pada model PLS sedikit lebih kecil dibandingkan pada model OLS. Oleh karena itu, dapat ditarik kesimpulan bahwa regresi PLS mampu mengatasi masalah multikolinearitas, namun nilai adjusted R² yang diperoleh masih cukup kecil. / In multicollinearity problems, independent variables that are highly correlated with other independent variables are removed, even though these variables have quite an impact on the dependent variable. Therefore, research to solve multicollinearity without the need to remove any variable is made. Partial Least Square (PLS) regression could be the solution to this problem; this method will be applied to analyze the foreign exchange rate towards Indonesian export-import value that has multicollinearity problems. The data used consists of 120 months of data from January 2012 until December 2021. This data will be split into train data and test data at a 90:10 ratio, then the data is transformed using the Cochrane-Orcutt method. PLS regression starts with significance tests to form PLS components, wherein the search for new components stops when all regression coefficients are not significant. Then, the regression between new components and dependent variable will be made to obtain the PLS model. The results of the PLS model succeed in solving the multicollinearity problem and fulfilled the four classical assumptions test. The results obtained will also be compared to results from Ordinary Least Square (OLS) with backward elimination method model. The Root Mean Square Error (RMSE) value obtained in the PLS model is smaller than the OLS model, while the adjusted R² value obtained in the PLS model is slightly smaller than the OLS model. Hence, it can be concluded that PLS regression can solve multicollinearity problems; however, the adjusted R² value obtained is relatively small.
Item Type: | Thesis (Bachelor) | ||||||||||||
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Creators: |
|
||||||||||||
Contributors: |
|
||||||||||||
Uncontrolled Keywords: | partial least square regression; multikolinearitas; kurs; ekspor; impor | ||||||||||||
Subjects: | Q Science > QA Mathematics | ||||||||||||
Divisions: | University Subject > Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics Current > Faculty/School - UPH Karawaci > Faculty of Science and Technology > Mathematics |
||||||||||||
Depositing User: | Christina Erica Sugianto Tejakusuma | ||||||||||||
Date Deposited: | 06 Jul 2022 02:35 | ||||||||||||
Last Modified: | 06 Jul 2022 02:35 | ||||||||||||
URI: | http://repository.uph.edu/id/eprint/48377 |
Actions (login required)
![]() |
View Item |