Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates

碩士 === 國立臺灣科技大學 === 營建工程系 === 107 === By providing a range of values rather than a point estimate, accurate interval forecasting is essential to the success of investment decisions in exchange rate markets. This study develops a sliding-window metaheuristic optimization for interval-valued time seri...

Full description

Bibliographic Details
Main Authors: Le Thi Thuy Linh, 李翠玲
Other Authors: Jui-Sheng Chou
Format: Others
Language:en_US
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/u7u4nb
id ndltd-TW-107NTUS5512003
record_format oai_dc
spelling ndltd-TW-107NTUS55120032019-05-16T01:40:46Z http://ndltd.ncl.edu.tw/handle/u7u4nb Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates Le Thi Thuy Linh 李翠玲 碩士 國立臺灣科技大學 營建工程系 107 By providing a range of values rather than a point estimate, accurate interval forecasting is essential to the success of investment decisions in exchange rate markets. This study develops a sliding-window metaheuristic optimization for interval-valued time series forecasting using multi-output least squares support vector regression (MLSSVR). The hyperparameters in MLSSVR are optimized using an accelerated particle swarm optimization algorithm to generate the best predictions and the fastest convergence. The proposed system has a graphical user interface developed in a computing environment and functions as a stand-alone application. The system is validated with stock price as well as exchange rates and outcomes are compared with previous results. Finally, the proposed interval time series prediction approach is tested in two case studies, one is the daily Australian dollar and Japanese yen rates (AUD/JPY) and the other involves US dollar and Canadian dollar rates (USD/CAD). The proposed model is promising for interval time series forecasting. Jui-Sheng Chou 周瑞生 2019 學位論文 ; thesis 106 en_US
collection NDLTD
language en_US
format Others
sources NDLTD
description 碩士 === 國立臺灣科技大學 === 營建工程系 === 107 === By providing a range of values rather than a point estimate, accurate interval forecasting is essential to the success of investment decisions in exchange rate markets. This study develops a sliding-window metaheuristic optimization for interval-valued time series forecasting using multi-output least squares support vector regression (MLSSVR). The hyperparameters in MLSSVR are optimized using an accelerated particle swarm optimization algorithm to generate the best predictions and the fastest convergence. The proposed system has a graphical user interface developed in a computing environment and functions as a stand-alone application. The system is validated with stock price as well as exchange rates and outcomes are compared with previous results. Finally, the proposed interval time series prediction approach is tested in two case studies, one is the daily Australian dollar and Japanese yen rates (AUD/JPY) and the other involves US dollar and Canadian dollar rates (USD/CAD). The proposed model is promising for interval time series forecasting.
author2 Jui-Sheng Chou
author_facet Jui-Sheng Chou
Le Thi Thuy Linh
李翠玲
author Le Thi Thuy Linh
李翠玲
spellingShingle Le Thi Thuy Linh
李翠玲
Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates
author_sort Le Thi Thuy Linh
title Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates
title_short Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates
title_full Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates
title_fullStr Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates
title_full_unstemmed Particle Swarm Optimized Multi-Output Support Vector Regression for Interval-Valued Forecasts of Exchange Rates
title_sort particle swarm optimized multi-output support vector regression for interval-valued forecasts of exchange rates
publishDate 2019
url http://ndltd.ncl.edu.tw/handle/u7u4nb
work_keys_str_mv AT lethithuylinh particleswarmoptimizedmultioutputsupportvectorregressionforintervalvaluedforecastsofexchangerates
AT lǐcuìlíng particleswarmoptimizedmultioutputsupportvectorregressionforintervalvaluedforecastsofexchangerates
_version_ 1719179128238768128