"Chapter 15 - Coronal Holes Detection Using Supervised Classification
Abstract
We demonstrate the use of machine learning algorithms in combination with segmentation techniques in order to distinguish coronal holes and filaments in solar extreme ultraviolet (EUV) images recorded by the Atmospheric Imaging Assembly onboard the Solar Dynamics Observatory. We used the Spatial Possibilistic Clustering Algorithm to prepare datasets of manually labeled coronal hole and filament channel regions present on the Sun during the time range 2010-16. By mapping the extracted regions from EUV observations onto Helioseismic and Magnetic Imager (HMI) line-of-sight magnetograms, we also include their magnetic characteristics. We computed average latitude, area, and shape measures from the segmented binary maps, as well as first-order and second-order texture statistics from the segmented regions in the EUV images and magnetograms. These attributes were used for data-mining investigations to identify the best rule for differentiating between coronal holes and filame!
nt channels, taking into account the imbalance in our dataset, which contains 1 filament channel for 15 coronal holes. We tested classifiers such as support vector machine (SVM), linear SVM, decision tree, k-nearest neighbors, as well as an ensemble classifier based on decision trees. The best performance in terms of true skill statistics is obtained with cost-sensitive learning, SVM classifiers, and when HMI attributes are included in the dataset.- Publication:
-
In: Machine Learning Techniques for Space Weather
- Pub Date:
- June 2018
- DOI:
- 10.1016/B978-0-12-811788-0.00015-9
- Bibcode:
- 2018mlts.book..365D
- Keywords:
-
- Solar wind;
- Coronal holes;
- Filament channels;
- Feature extraction;
- Supervised classification;
- Textural features