Exchange rate prediction using computational intelligence
Foreign exchange (FOREX) market has become a trend in the past few years which has been heavily covered by a number of researchers in order to come out with the reliable price forecasting. The thesis covered some methods from the field of computational intelligence to predict the exchange rate, comp...
Saved in:
Main Author: | |
---|---|
Other Authors: | |
Format: | Final Year Project |
Language: | English |
Published: |
2017
|
Subjects: | |
Online Access: | http://hdl.handle.net/10356/71192 |
Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
Institution: | Nanyang Technological University |
Language: | English |
id |
sg-ntu-dr.10356-71192 |
---|---|
record_format |
dspace |
spelling |
sg-ntu-dr.10356-711922023-07-07T16:04:23Z Exchange rate prediction using computational intelligence Kurniawan, Nikolas Ponnuthurai Nagaratnam Suganthan School of Electrical and Electronic Engineering DRNTU::Engineering::Electrical and electronic engineering Foreign exchange (FOREX) market has become a trend in the past few years which has been heavily covered by a number of researchers in order to come out with the reliable price forecasting. The thesis covered some methods from the field of computational intelligence to predict the exchange rate, comprises of a number of linear methods and machine learning techniques. Besides, a couple of data adjustment methods and features selection are applied to facilitate the system to produce an accurate prediction, which is found very critical to the system performance. It is found that the data set that has been decomposed based on the frequency performs relatively better than the untreated data. The best performance in terms of RMSE is the RVFL model with 1000 neurons and 6 input features when it is used with the decomposed price data. When the best algorithms are applied to all of the currency pairs used in the project, most of them have an RMSE value of less than 0.1 and direction accuracy around 80% given that all the previous values are known. The four best algorithms are used for the hybrid algorithm development which truncates the direction error even further than any single method with the accuracy of almost 90%. Bachelor of Engineering 2017-05-15T07:13:21Z 2017-05-15T07:13:21Z 2017 Final Year Project (FYP) http://hdl.handle.net/10356/71192 en Nanyang Technological University 83 p. application/pdf |
institution |
Nanyang Technological University |
building |
NTU Library |
continent |
Asia |
country |
Singapore Singapore |
content_provider |
NTU Library |
collection |
DR-NTU |
language |
English |
topic |
DRNTU::Engineering::Electrical and electronic engineering |
spellingShingle |
DRNTU::Engineering::Electrical and electronic engineering Kurniawan, Nikolas Exchange rate prediction using computational intelligence |
description |
Foreign exchange (FOREX) market has become a trend in the past few years which has been heavily covered by a number of researchers in order to come out with the reliable price forecasting. The thesis covered some methods from the field of computational intelligence to predict the exchange rate, comprises of a number of linear methods and machine learning techniques. Besides, a couple of data adjustment methods and features selection are applied to facilitate the system to produce an accurate prediction, which is found very critical to the system performance. It is found that the data set that has been decomposed based on the frequency performs relatively better than the untreated data. The best performance in terms of RMSE is the RVFL model with 1000 neurons and 6 input features when it is used with the decomposed price data. When the best algorithms are applied to all of the currency pairs used in the project, most of them have an RMSE value of less than 0.1 and direction accuracy around 80% given that all the previous values are known. The four best algorithms are used for the hybrid algorithm development which truncates the direction error even further than any single method with the accuracy of almost 90%. |
author2 |
Ponnuthurai Nagaratnam Suganthan |
author_facet |
Ponnuthurai Nagaratnam Suganthan Kurniawan, Nikolas |
format |
Final Year Project |
author |
Kurniawan, Nikolas |
author_sort |
Kurniawan, Nikolas |
title |
Exchange rate prediction using computational intelligence |
title_short |
Exchange rate prediction using computational intelligence |
title_full |
Exchange rate prediction using computational intelligence |
title_fullStr |
Exchange rate prediction using computational intelligence |
title_full_unstemmed |
Exchange rate prediction using computational intelligence |
title_sort |
exchange rate prediction using computational intelligence |
publishDate |
2017 |
url |
http://hdl.handle.net/10356/71192 |
_version_ |
1772828582721093632 |