Tyas, Sasmi Hidayatul Yulianing (2015) Implementasi Peramalan Harga Emas Berbasis Web Yang Didasarkan Pada Hasil Analisis Pengaruh Harga Minyak, Nilai Tukar Mata Uang, Dan Indeks Standard & Poor. Undergraduate thesis, Institut Technology Sepuluh Nopember.
Preview |
Text
5211100073-Undergraduate Thesis.pdf - Published Version Download (2MB) | Preview |
Abstract
Emas merupakan salah satu objek investasi yang paling diminati.
Hal ini ditunjukkan oleh Indeks Dow Jones pada tahun 2014 yang
mengindikasikan bahwa 12% dari keseluruhan investasi melibatkan
objek emas. Di lain pihak, sejumlah penelitian menunjukkan
bahwa harga emas dipengaruhi oleh beberapa variabel lain
seperti harga minyak (Europe Brent Spot Price FOB dan Chusing
OK WTI Spot Price FOB), nilai tukar mata uang USD terhadap
mata uang lainnya (AUD, GBP, CAD, INR, dan EUR), dan indeks
Standard & Poor (indeks SP). Dengan demikian, aplikasi peramalan
yang didasarkan pada hasil analisis pengaruh beberapa
variabel tersebut terhadap harga emas sangat diperlukan untuk
membantu para investor dalam memprediksi keuntungan yang
dapat diperoleh dalam berinvestasi emas.
Dalam Tugas Akhir ini, analisis pengaruh variabel independen
(harga minyak, nilai tukar mata uang USD terhadap beberapa
mata uang lainnya), dan indeks SP terhadap harga emas sebagai
variabel dependen dilakukan menggunakan metode regresi linier
berganda. Analisis dilakukan untuk beberapa komposisi perbandingan
data pelatihan dan data pengujian menggunakan data
vi
yang diperoleh dari berbagai sumber dalam rentang waktu mulai
tanggal 2 Januari 2013 sampai dengan tanggal 31 Desember
2014. Hasil perhitungan regresi kemudian digunakan untuk melakukan
analisis pengaruh dan hubungan antar variabel melalui
dua jenis pengujian, yaitu uji goodness of fit (yang terdiri dari
penentuan koefisien determinasi, uji t, dan uji f) dan uji asumsi
klasik (yang terdiri dari uji normalitas, uji otokorelasi, uji multikolinieritas,
dan uji heterokesdatisitas). Aplikasi peramalan berbasis
Web kemudian dibangun berdasarkan hasil analisis yang
memberikan nilai mean absolute percentage error (MAPE) terendah.
Hasil analisis menunjukkan bahwa pada dasarnya semua
variabel independen (harga minyak, nilai tukar mata uang, dan
indeks SP) mempunyai pengaruh yang cukup signifikan terhadap
harga emas. Selain itu, hasil analisis menunjukkan bahwa walaupun
variabel harga minyak WTI terindikasi mempunyai pengaruh
yang lemah terhadap harga emas, tetapi keberadaannya dapat
mempengaruhi kualitas hasil peramalan, sehingga variabel tersebut
tidak dapat diabaikan. Dari hasil analisis tersebut, aplikasi
peramalan harga emas berbasis Web kemudian diimplementasikan
berdasarkan komposisi perbandingan data pelatihan dan
data pengujian sebesar 90%:10% yang menghasilkan nilai
MAPE terendah sebesar 14,73%.
=======================================================================================================
Gold is becoming one of the most attractive investment object.
This is shown by Dow Jones Indexes in 2014 which indicated that
gold usage in investment is 12% of the total investment objects.
On the other hand, some studies found that the gold price is
influenced by some other variables, such as oil prices (Europe
Brent Spot Price FOB and Chusing OK WTI Spot Price FOB),
exchange rates of USD against other currencies (AUD, GBP,
CAD, INR , and EUR), and the Standard & Poor's (SP) indexes.
Therefore, a forecasting application based on analytical results of
the influence of several variables on the price of gold is needed to
assist investors in predicting the profits to be gained in gold
investment.
In this final project, the analysis of the influence of the independent
variables (oil prices, exchange rates of USD against
some other currencies, and the SP indexes) on the price of gold
as the dependent variable was performed using multiple linear
regression method. The analysis was performed at some ratio of
training and testing data using data that was obtained from various
sources ranging from January 2, 2013 until December 31,
2014. Results of the regression computation procedure are then
used to analyze the influence and relationship among variables
through two types of testing; i.e., the goodness of fit test consisting
of the setting of the coefficient determination, t and f tests
viii
as well as classical assumptions consisting of normality, autocorrelation,
multicolinearity, and heterocesdatisity tests. Finally,
a web-based forecasting application is then built based on the
analysis results having the lowest mean absolute percentage
error (MAPE).
The analysis results showed that essentially all independent variables
(i.e., oil prices, exchange rates, and SP indexes) have a significant
influence on the gold price. Although, the analysis results
showed that WTI oil price variable has indicated a weak influence
on the gold price, but its presence still influences the quality
of forecasting results, and hence this variable can not be excluded.
A web-based application of gold price forecasting is then implemented
based on the composition ratio of 90%:10% between
training and testing data that produces the lowest MAPE value of
14.73%.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSI 005.1 Tya i |
| Uncontrolled Keywords: | aplikasi peramalan, pengaruh, investasi, harga emas, harga minyak, nilai tukar mata uang, indeks SP, metode regresi linier berganda. |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.888 Web sites--Design. Web site development. |
| Divisions: | Faculty of Information Technology > Information System > 57201-(S1) Undergraduate Thesis |
| Depositing User: | Mr. Tondo Indra Nyata |
| Date Deposited: | 29 Oct 2019 02:32 |
| Last Modified: | 29 Oct 2019 02:32 |
| URI: | http://repository.its.ac.id/id/eprint/71474 |
Actions (login required)
 |
View Item |