Forecast daily indices of solar activity, F10.7, using support vector regression method

Authors:

Abstract
The 10.7cm solar radio flux (F10.7), the value of the solar radio emission flux density at wavelength 10.7 cm, is a use- ful index of solar activity as a proxy for solar extreme ultra-violet (EUV) radiation. It is meaningful and important to predict F10.7 values accurately for both long-term (months-years) and short- term (days) forecasting,which are often used as inputs in the space weather models. This study applies a novel neural network tech- nique, supportvectorregression(SVR), to forecastingdaily values of F10.7.The aim of this study is to examinethe feasibility of SVR in short-term F10.7 forecasting. The approach based on SVR re- duces the dimension of feature space in training process by using kernel-based learning algorithm. Thus, the calculating complexity becomes lower and a small amount of training data will be suffi- cient. The time series of F10.7 during 2002 to 2006 are employed as the data sets. The performance of the approach is estimated by calculating the norm mean square error (NMSE) and mean abso- lute percentage error (MAPE). It is shown that the approach can perform well by using fewer training data points than the tradi- tional neural network.

This publication has 8 references indexed in Scilit: