Mathematics – Probability
Scientific paper
May 2007
adsabs.harvard.edu/cgi-bin/nph-data_query?bibcode=2007aas...210.6805f&link_type=abstract
American Astronomical Society Meeting 210, #68.05; Bulletin of the American Astronomical Society, Vol. 39, p.176
Mathematics
Probability
Scientific paper
Automated forecasting the onset of solar flares from the analysis of photospheric magnetogram data remains an essential and challenging task. In the present study, we present a novel kernel-based regression method for predicting the probability distribution of flare index of an active region. The target variable, soft X-ray flare index, quantifies the flare productivity of an active region within the chosen time window. The predictor vector of an active region includes several magnetic parameters which are derivable from the MDI line-of-sight magnetograms (e.g., total unsigned magnetic flux, the length of the magnetic neutral lines with strong magnetic gradient, etc). By applying kernel functions, the predictor vectors are implicitly mapped into high dimensional feature space which is more informative than the original input space. Then the regression analysis is conducted in this feature space. Compared to the conventional statistical regression analysis, kernel-based methods have shown great advantages. Details of the method and data analysis procedure are first described in the paper. We then applied the method to a large sample dataset of active regions (NOAA 7961 - 10933). The experimental results are presented, showing that our method is of practical significance in automated flare forecasting.
Fu Gang
Jing Ji-liang
Shih Frank Y.
Song Hui-chao
Wang Hai-Hong
No associations
LandOfFree
Solar Flare Prediction Using Kernel-based Regression Analysis does not yet have a rating. At this time, there are no reviews or comments for this scientific paper.
If you have personal experience with Solar Flare Prediction Using Kernel-based Regression Analysis, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Solar Flare Prediction Using Kernel-based Regression Analysis will most certainly appreciate the feedback.
Profile ID: LFWR-SCP-O-1031309