Rosyidah, Anis Latif (2015) Pembuatan Aplikasi Prediksi Harga Saham Berbasis Web Menggunakan Fuzzy Time Series: Studi Kasus Di Bursa Efek Indonesia - Development Of A Web-Based Stock Price Prediction Application Using Fuzzy Time Series : A Case Study At Indonesia Stock Exchange. Undergraduate thesis, Institut Teknologi Sepuluh Nopember.
Preview |
Text
5211100090-Undergraduate Thesis.pdf - Published Version Download (2MB) | Preview |
Preview |
Text
5211100090-Paper.pdf - Published Version Download (346kB) | Preview |
Preview |
Text
5211100090-Presentation.pdf - Published Version Download (608kB) | Preview |
Abstract
Bursa Efek Indonesia (BEI) merupakan salah satu tempat
yang paling menarik bagi para investor untuk melakukan jualbeli
saham di Indonesia. Indeks harga saham LQ45 merupakan
salah satu index harga saham yang dilirik oleh para investor
karena di dalamnya tergabung 45 perusahaan besar di
Indonesia yang dianggap memiliki prospek bisnis yang bagus.
Bagi para investor prediksi pergerakan indeks harga saham
menjadi sangat penting untuk membantu mereka dalam memutuskan
waktu yang tepat untuk menjual atau membeli saham.
Oleh karenanya, aplikasi yang mampu melakukan prediksi
indeks harga saham dengan akurat dan dapat digunakan
dengan mudah oleh masyarakat umum menjadi penting untuk
disediakan.
Tugas Akhir ini barkaitan dengan pembuatan aplikasi peramalan
harga saham berbasis Web yang didasarkan pada
metode peramalan data runtut waktu berbasis logika fuzzy
(fuzzy time series). Penggunaan peramalan harga saham berbasis
logika fuzzy dilakukan untuk mengatasi keterbatasan
metode peramalan runtut waktu biasa yang tidak mampu
menerjemahkan nilai lingusitik terhadap perubahan kenaikan
atau penurunan data harga saham. Model peramalan harga
saham berbasis logika fuzzy dibangun dengan dengan
menggunakan data harian harga saham LQ45 mulai dari tahun 2011 sampai dengan tahun 2014. Sebelum diimplementasikan
menjadi sebuah aplikasi berbasis Web,
berbagai uji coba kinerja hasil peramalan dilakukan untuk
menjamin keandalan dari model peramalan yang dibuat. Pengujian
tersebut antara lain meliputi: uji penentuan komposisi
prbandingan data pelatihan dan data validasi, uji penentuan
jumlah interval data untuk memperoleh himpunan fuzzy yang
optimal, uji sensitivitas penambahan data baru, dan uji perbandingan
dengan model peramalan berbasis single moving
average.
Hasil uji coba kinerja model peramalan yang telah berhasil
diimplemetasikan dalam sebuah aplikasi berbasis Web
menghasilkan nilai mean absolute percentage error (MAPE)
sebesar 0,36% untuk model peramalan dengan komposisi
perbandingan data pelatihan dan validasi sebesar 75%:25%
dan jumlah interval data sebanyak 11. Hasil ini jauh lebih
baik dibandingkan dengan hasil peramalan menggunakan
metode peramalan runtut waktu baisa (single moving average)
yang hanya memberikan nilai MAPE sebesar 27%. Selain itu,
hasil uji sensitvitas penambahan data menunjukkan bahwa
model peramalan runtut waktu berbasis logika fuzzy mampu
melakukan peramalan dengan kinerja yang konsisten walaupun
penambahan data baru dilakukan untuk periode tahunan.
========================================================================================================================
Indonesia Stock Exchange (IDX) is one of the most interesting
places for investors to buy and sell stocks in Indonesia. LQ45
stock price index is becoming the one remarked by investors
since it incorporated 45 large companies in Indonesia that are
considered having good business prospects. For investors,
forecast stock price index movement is very important to help
them decide the right time to sell or buy stocks. Therefore, an
application which is capable of predicting stock price accurately
and easy to use becomes necessary to be provided.
This final project is concerned with a web-based stocks price
prediction application development using the forecasting
model based on fuzzy time series. In this forecasting model,
fuzzy logic is used to overcome the limitations of conventional
time series forecasting methods that are not capable of
translating the linguistic values agaist the increasing or
decreasing of stocks data. The stock price forecasting model
based on fuzzy logic is developed using daily data of LQ45
stock prices from 2011 until 2014. Before being implemented
into a web-based application, a number of performance tests
were performed on the forecasting model in order to ensure
the reliability of the application being developed. These tests
include: determining the composisiton of pelatihan and validation data, determining the number of data intervals
testing to obtain an optimal fuzzy set, the sensitivity tests due
to the addition of new data, and comparison test with a single
moving average forcasting model.
Results of the forecasting model performance tests that have
been successfully implemented in a web-based application
produced mean absolute percentage error (MAPE) value of
0.36% for the forecasting model with the composition of the
pelatihan and validation data by 75%:25% and with the
number of data intervals of 11. The MAPE value of this model
is much better than that produces by the conventional time
series forecasting model (i.e., single moving average) which
only gives the MAPE value of 27%. In addition to this, the
result of sensitivity test due to addition of new data showed
that the time series forecasting model based on fuzzy logic is
capable of producing consistent performance although the
addition of new data was made annually.
| Item Type: | Thesis (Undergraduate) |
|---|---|
| Additional Information: | RSSI 006.7 Ros p |
| Uncontrolled Keywords: | prediksi, harga saham, Bursa Efek Indonesia, fuzzy time series, LQ45, aplikasi berbasis Web, prediction, stock price, Indonesia Stock Exchange, fuzzy time series, LQ45, Web-based application |
| Subjects: | T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.888 Web sites--Design. Web site development. |
| Divisions: | Faculty of Information Technology > Information System |
| Depositing User: | ansi aflacha |
| Date Deposited: | 22 Nov 2019 06:29 |
| Last Modified: | 22 Nov 2019 06:29 |
| URI: | http://repository.its.ac.id/id/eprint/71971 |
Actions (login required)
 |
View Item |