2009 Sixth International Conference on Fuzzy Systems and Knowledge Discovery A Thunderstorm Forecast Model Based on Weighted SVM and Data Field Tianjin, China August 14-August 16 ISBN: 978-0-7695-3735-1
DOI Bookmark: http://doi.ieeecomputersociety.org/10.1109/FSKD.2009.194
To solve imbalance problem of datasets in thunderstorm forecast, this paper introduced the concept of data field and proposed a resampling method based on potential value which is combined with the weighted Support Vector Machine [12-14] (SVM) to set up a new thunderstorm forecast model. Moreover we assessed the forecast model with a comprehensive assessment method based on imbalance measure and meteorological score. The experimental results showed that the model effectively controlled the adverse impact of unbalanced datasets to thunderstorm forecast. By the assessment of comprehensive assessment method, the results proved that the model is not only effective in dealing with the imbalance datasets, but also more practical in weather forecast.
Index Terms:
SVM, unbalanced datasets, data field, g-means, CSI
Citation:
Wei Fan, Jie Ma, He Zhu, "A Thunderstorm Forecast Model Based on Weighted SVM and Data Field," fskd, vol. 5, pp.160-164, 2009 Sixth International Conference on Fuzzy Systems and Knowledge Discovery, 2009 Usage of this product signifies your acceptance of the Terms of Use. | |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Digg
Furl
Spurl
Blink
Simpy
Google
Del.icio.us
Y!MyWeb