Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models
Approaches for predicting financial markets, including conventional statistical methods and recent deep learning methods, have been investigated in many studies. However, financial time series data (e.g., daily stock market index) contain noises that prevent stable predictive model learning. Using t...
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
IEEE
2021-01-01
|
| Series: | IEEE Access |
| Subjects: | |
| Online Access: | https://ieeexplore.ieee.org/document/9446858/ |
| id |
doaj-f31e10a6c6494a98a4540c25b8c17bde |
|---|---|
| record_format |
Article |
| spelling |
doaj-f31e10a6c6494a98a4540c25b8c17bde2021-06-14T23:00:21ZengIEEEIEEE Access2169-35362021-01-019837868379610.1109/ACCESS.2021.30865379446858Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning ModelsDonghwan Song0Adrian Matias Chung Baek1https://orcid.org/0000-0002-4372-1866Namhun Kim2https://orcid.org/0000-0003-4429-2191Department of System Design and Control Engineering, Ulsan National Institute of Science and Technology, Ulsan, South KoreaDepartment of Mechanical Engineering, Ulsan National Institute of Science and Technology, Ulsan, South KoreaDepartment of Mechanical Engineering, Ulsan National Institute of Science and Technology, Ulsan, South KoreaApproaches for predicting financial markets, including conventional statistical methods and recent deep learning methods, have been investigated in many studies. However, financial time series data (e.g., daily stock market index) contain noises that prevent stable predictive model learning. Using these noised data in predictions results in performance deterioration and time lag. This study proposes padding-based Fourier transform denoising (P-FTD) that eliminates the noise waveform in the frequency domain of financial time series data and solves the problem of data divergence at both ends when restoring to the original time series. Experiments were conducted to predict the closing prices of S&P500, SSE, and KOSPI by applying data, from which noise was removed by P-FTD, to different deep learning models based on time series. Results show that the combination of the deep learning models and the proposed denoising technique not only outperforms the basic models in terms of predictive performance but also mitigates the time lag problem.https://ieeexplore.ieee.org/document/9446858/Deep learningdenoising frameworkFourier transformstock index predictiontime series |
| collection |
DOAJ |
| language |
English |
| format |
Article |
| sources |
DOAJ |
| author |
Donghwan Song Adrian Matias Chung Baek Namhun Kim |
| spellingShingle |
Donghwan Song Adrian Matias Chung Baek Namhun Kim Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models IEEE Access Deep learning denoising framework Fourier transform stock index prediction time series |
| author_facet |
Donghwan Song Adrian Matias Chung Baek Namhun Kim |
| author_sort |
Donghwan Song |
| title |
Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models |
| title_short |
Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models |
| title_full |
Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models |
| title_fullStr |
Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models |
| title_full_unstemmed |
Forecasting Stock Market Indices Using Padding-Based Fourier Transform Denoising and Time Series Deep Learning Models |
| title_sort |
forecasting stock market indices using padding-based fourier transform denoising and time series deep learning models |
| publisher |
IEEE |
| series |
IEEE Access |
| issn |
2169-3536 |
| publishDate |
2021-01-01 |
| description |
Approaches for predicting financial markets, including conventional statistical methods and recent deep learning methods, have been investigated in many studies. However, financial time series data (e.g., daily stock market index) contain noises that prevent stable predictive model learning. Using these noised data in predictions results in performance deterioration and time lag. This study proposes padding-based Fourier transform denoising (P-FTD) that eliminates the noise waveform in the frequency domain of financial time series data and solves the problem of data divergence at both ends when restoring to the original time series. Experiments were conducted to predict the closing prices of S&P500, SSE, and KOSPI by applying data, from which noise was removed by P-FTD, to different deep learning models based on time series. Results show that the combination of the deep learning models and the proposed denoising technique not only outperforms the basic models in terms of predictive performance but also mitigates the time lag problem. |
| topic |
Deep learning denoising framework Fourier transform stock index prediction time series |
| url |
https://ieeexplore.ieee.org/document/9446858/ |
| work_keys_str_mv |
AT donghwansong forecastingstockmarketindicesusingpaddingbasedfouriertransformdenoisingandtimeseriesdeeplearningmodels AT adrianmatiaschungbaek forecastingstockmarketindicesusingpaddingbasedfouriertransformdenoisingandtimeseriesdeeplearningmodels AT namhunkim forecastingstockmarketindicesusingpaddingbasedfouriertransformdenoisingandtimeseriesdeeplearningmodels |
| _version_ |
1721377868129239040 |