A sequential data mining method for modelling solar magnetic cycles

摘要: We propose an adaptive data-driven approach to modelling solar magnetic activity cyclesbased on a sequential link between unsupervised and supervised modelling. Monthly sunspot numbers spanning over hundreds of years --- from the mid-18th century first quarter 2012 - obtained Royal Greenwich Observatory provide reliable source training validation sets.An indicator variable is used generate class labels internal parameters which are separate high low cycles. Our results show that by maximising data-dependent using them as inputs support vector machine model we obtain comparatively more robust predictions. Finally, demonstrate how method can be adapted other applications.