Pattern recognition techniques in Polarimetry

Sparsity is a property of data by which it can be represented using a small number of patterns. It is the key concept behind an evergrowing list of mathematical techniques for handling data and recover from it signals or information in conditions previously thought impossible. The application of tho...

Full description

Saved in:
Bibliographic Details
Published in:Proceedings of the International Astronomical Union 2014-12, Vol.10 (S305), p.207-215
Main Author: Ariste, Arturo López
Format: Article
Language:English
Subjects:
Astronomy
Compressed
Contributed Papers
Data analysis
Lists
Magnetic fields
Mathematical analysis
Pattern recognition
Polarimetry
Solar prominences
Stars & galaxies
Vectors (mathematics)
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Sparsity is a property of data by which it can be represented using a small number of patterns. It is the key concept behind an evergrowing list of mathematical techniques for handling data and recover from it signals or information in conditions previously thought impossible. The application of those techniques to spectropolarimetric data is relatively straightforward. We present three examples of such application: the use of Principal Component Analysis to invert the magnetic field in solar prominences from spectropolarimetry of the He D3 line, the removal of fringes from spectropolarimetric data with Relevance Vector Machines, and the retrieval of high resolution spectra from low resolution data with Compressed Sensing.
ISSN:1743-9213
1743-9221
DOI:10.1017/S1743921315004792