Exploring X-ray variability with unsupervised machine learning
نویسندگان
چکیده
XMM-Newton provides unprecedented insight into the X-ray Universe, recording variability information for hundreds of thousands sources. Manually searching interesting patterns in light curves is impractical, requiring an automated data-mining approach characterization Straightforward fitting temporal models to not a sure way identify them, especially with noisy data. We used unsupervised machine learning distill large data set light-curve parameters, revealing its clustering structure preparation anomaly detection and subsequent searches specific source behaviors (e.g., flares, eclipses). Self-organizing maps (SOMs) achieve dimensionality reduction within single framework. They are type artificial neural network trained approximate two-dimensional grid discrete interconnected units, which can later be visualized on plane. our SOM temporal-only parameters computed from more than 100,000 detections EXTraS catalog. The resulting map reveals that about 2500 most variable sources clustered based characteristics. find distinctive regions associated eclipses, dips, linear curves, others. Each group contains appear similar by eye. out handful further study. condensed view dataset provided SOMs allowed us groups sources, speeding up manual orders magnitude. Our method also highlights problems simple mitigate them extent. This will crucial fully exploiting high volume expected upcoming surveys, may help interpreting supervised classification models.
منابع مشابه
Machine Learning for NLP: Unsupervised learning techniques
• So far we have seen supervised learning (of classification): – learning based on a training set where labelling of instances represents the target (categorisation) function – classifier implements an approximation of the target funtion – outcome: a classification decision • Unsupervised learning: – learning based on unannotated instances; – outcome: a grouping of objects (instances and groups...
متن کاملExploring the X-ray spectral variability of MCG–6-30-15 with XMM-Newton
We present a study of the spectral variability of the Seyfert I galaxy MCG–6-3015 based on the two long XMM-Newton observations from 2000 and 2001. The X– ray spectrum and variability properties of the 2001 data have previously been well described with a two-component model consisting of a variable power law and a much less variable reflection component, containing a broad relativistic iron lin...
متن کاملTackling the X-ray cargo inspection challenge using machine learning
The current infrastructure for non-intrusive inspection of cargo containers cannot accommodate exploding commerce volumes and increasingly stringent regulations. There is a pressing need to develop methods to automate parts of the inspection workflow, enabling expert operators to focus on a manageable number of high-risk images. To tackle this challenge, we developed a modular framework for aut...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Astronomy and Astrophysics
سال: 2022
ISSN: ['0004-6361', '1432-0746']
DOI: https://doi.org/10.1051/0004-6361/202142444