Deep learning for the automatic detection of interplanetary coronal mass ejections

Interplanetary coronal mass ejections (ICMEs) are one of the main drivers for space weather disturbances. In the past, their detection in solar wind in situ data was subject to time consuming and often biased expert labeling. Within this thesis, we reimplement an existing method based on deep learning for the automatic detection of ICMEs and test it on additional data. Furthermore, we use an alternative validation method focusing on generalization, as conventional within the machine learning community, to avoid overfitting and predict the performance on new unseen data more accurately. This method is then compared to our own method which was developed within this thesis. Finally, we discuss their respective advantages and give prospects to possible future applications. (author)