Luo, Yaqing (2021) Wavelet neural network model with time-frequency analysis for accurate share prices prediction. In: Arai, Kohei, (ed.) Intelligent Computing - Proceedings of the 2021 Computing Conference: Volume 3. Lecture Notes in Networks and Systems,3. Springer Science and Business Media Deutschland GmbH, Cham, CH, 286 - 297. ISBN 9783030801281
Full text not available from this repository.Abstract
Due to the large amounts of risks and potential financial benefits involved, the ability to achieve accurate prediction on stock market prices is of great interest to investors. However, the non-stationarity, high level of volatility, frequent fluctuations and stochastic properties that the data possesses, have made it difficult to accurately predict share prices, even by recently developed deep learning methods. This can be attributed to the outputs trained that are not responsive enough to capture the rapid adjustments in real data, hence affecting prediction accuracy. To solve these difficulties, this paper proposes a wavelet neural network model by using Gaussian wavelet as activation function and decomposing share prices data into finer precision with wavelet to account for the sensitivity, and further optimising the neural network mapping and learning process with detailed time-frequency analysis of outputs, leading to higher prediction accuracy and faster learning speed. The proposed model with two training processes has been validated using the dataset from London stock market, and the results have demonstrated that the wavelet neural network model-based predictions are distinctly superior to that of current deep learning methods, which corresponds to a significant reduction in mean squared error.
| Item Type: | Book Section |
|---|---|
| Official URL: | https://link.springer.com/book/10.1007/978-3-030-8... |
| Additional Information: | © 2021 The Author, under exclusive license to Springer Nature Switzerland AG. |
| Divisions: | Mathematics |
| Subjects: | Q Science > QA Mathematics |
| Date Deposited: | 30 Mar 2022 10:24 |
| Last Modified: | 09 Nov 2024 17:24 |
| URI: | http://eprints.lse.ac.uk/id/eprint/114529 |
Actions (login required)
 |
View Item |
