Hermine Landt

Science Collaboration(s): 
Informatics and Statistics (ISSC)

I use the reverberation echo-mapping technique in order to study the inner structure of AGN, in particular the centrally obscuring dust emitting at infrared wavelengths. Such work involves modelling of optical photometric and spectroscopic light-curves, which are subsequently smoothed and shifted in order to determine the time delay between time series observed in different frequency bands. Currently preferred models are based on Gaussian Processes, which we have used successfully and extensively. Now I am interested to explore the application of other efficient Machine Learning algorithms for time-series analysis, such as, eg., Recurrent Neural Networks (RNNs) and models based on Maximum Entropy.