A New Paradigm in X-ray Spectral Fitting using Recurrent Neural Networks
Using a class of neural networks known as a Recurrent Inference Machine, we have successfully deconvolved the source’s intrinsic spectrum from the instrumental response function for the first time.
A New Paradigm in X-ray Spectral Fitting using Recurrent Neural Networks
A New Paradigm in X-ray Spectral Fitting
This short video explains recent advances in deconvolving X-ray spectra using recurrent neural network architectures. Deconvolving X-ray spectra allows us to access, for the first time, the intrinsic spectrum of the source — in this case, the intracluster medium [1]. With the intrinsic spectra, we are able to stack data, explore the transiency of X-ray sources, and unlock the use of novel machine learning algorithms.
Creative Commons Attribution 4.0 International License (CC-BY 4.0)
