Ouarbya, Lahcen; Mirikitani, Derrick Takeshi and Martin, Eamonn. 2012. The use of sequential recurrent neural filters in forecasting the Dst index for the strong magnetic storm of autumn 2003. Applied Mathematics Letters, 25(10), pp. 1361-1366. ISSN 0893-9659 [Article]

No full text available

Text Seek_main_paper.pdf - Published Version Permissions: Administrator Access Only Download (347kB)

Abstract or Description

Neural based geomagnetic forecasting literature has heavily relied upon non-sequential algorithms for estimation of model parameters. This paper proposes sequential Bayesian recurrent neural filters for online forecasting of the Dst index. Online updating of the RNN parameters allows for newly arrived observations to be included into the model. The online RNN filters are compared to two (non-sequentially trained) models on a severe double storm that has so far been difficult to forecast. It is shown that the proposed models can significantly reduce forecast errors over non-sequentially trained recurrent neural models.

Item Type:	Article
Identification Number (DOI):	https://doi.org/10.1016/j.aml.2011.12.002
Keywords:	Forecasting magnetospheric disturbances, Recurrent neural networks
Departments, Centres and Research Units:	Computing
Dates:	Date Event 5 December 2011 Accepted 17 December 2011 Published Online October 2012 Published
Item ID:	30982
Date Deposited:	11 Jan 2022 11:43
Last Modified:	11 Jan 2022 11:43
Peer Reviewed:	Yes, this version has been peer-reviewed.
URI:	https://research.gold.ac.uk/id/eprint/30982

View statistics for this item...

Edit Record (login required)