Improving Time Series Models Prediction Based On Empirical Mode Decomposition Using Stock Market Data

Hossain, Mohammad Raquibul (2021) Improving Time Series Models Prediction Based On Empirical Mode Decomposition Using Stock Market Data. PhD thesis, Universiti Sains Malaysia..

Download (357kB) | Preview


Time series analysis and prediction is a very important and active research area. In this age of profuse data generation, proper use of available data has become crucial in forecasting and decision making. This thesis presents the research study involving the development of five advanced forecasting methods and their experimentation on twelve stock price time series datasets. Traditional forecasting methods have limitations in forecasting potentiality due to their linearity and stationarity assumptions on the datasets. However, real life data including stock price data have sophisticated features and patterns encompassing nonlinearity and non-stationarity. Therefore, there is the research scope to search for better methods to improve forecast accuracy obtainable from the traditional methods by applying advanced approaches. Empirical mode decomposition (EMD), a very essentially important part of Hilbert-Huang transforms (HHT) is a very adaptive decomposition algorithm to view data from granular and different time scales. Being a robust analysing tool in signal processing, EMD has been widely applied in other fields including economics and finance. However, there are still scopes in improving the forecast accuracy of nonlinear nonstationary financial time series using EMD and other forecasting methods. From such relevant hypotheses, this study was followed by three research objectives. Five EMD-based methods were developed on these objectives

Item Type: Thesis (PhD)
Subjects: Q Science > QA Mathematics > QA1 Mathematics (General)
Divisions: Pusat Pengajian Sains Matematik (School of Mathematical Sciences) > Thesis
Depositing User: Mr Noor Azizan Abu Hashim
Date Deposited: 30 Jun 2022 17:16
Last Modified: 30 Jun 2022 17:16

Actions (login required)

View Item View Item