Short-term prediction of solar flare activity using multiple regression methods was considered. The variables describing active regions the given day were used to predict the flare activity on the next day. Two groups of observational data covering the years 1988 and 1989 were dealt with. Some variants of the distance-based regression as proposed by Cuadras and Arenas (1990) appeared to be superior to the ordinary least squares method by describing more accurately the data sets under consideration.
@article{urn:eudml:doc:40177,
title = {Distance-based regression in prediction of solar flare activity.},
journal = {Q\"uestii\'o},
volume = {18},
year = {1994},
pages = {7-38},
mrnumber = {MR1288940},
zbl = {1167.85303},
language = {en},
url = {http://dml.mathdoc.fr/item/urn:eudml:doc:40177}
}
Bartkowiak, Anna; Jakimiec, Maria. Distance-based regression in prediction of solar flare activity.. Qüestiió, Tome 18 (1994) pp. 7-38. http://gdmltest.u-ga.fr/item/urn:eudml:doc:40177/