For the purpose of fast methods to handle huge amount of data
coming from future solar spectropolarmeter, the statistical machine learning techniques based on Mercer?s kernel have been applied to the inversion of the photospheric magnetic fields from the polarimetric data. In particular, the Regu-
larized Neural Network and the Support Vector Machine have been tested for the data from HMI (Helioseismic and Magnetic Imager ) on SDO (Solar Dynamics Observatory).
Authors: Fei Teng
Publication Status: accepted by Solar Phys.
Last Modified: 2015-09-14 16:10