PyFIRSTClassifier 2.0.2

Automated morphological classification of Compact and Extended radio sources using Deep Convolutional Neural Networks

FIRST Classifier: Compact and Extended Radio Galaxies Classification using Deep Convolutional Neural Networks

The FIRST Classifier is a tool for the automated morphological classification of radio sources based on data from the FIRST radio survey. This system leverages a trained Deep Convolutional Neural Network to classify radio galaxies into Compact, BENT, FRI, and FRII categories with high accuracy. It can predict the morphological class of single or multiple sources. - Accuracy: 97% - Recall: Compact (98%), BENT (100%), FRI (98%), FRII (93%)

How to use:

   from PyFIRSTClassifier import FIRSTClassifier

   classifier = FIRSTClassifier.Classifiers()

   # Example for single source classification
   ra = 223.47337
   dec = 26.80928

   # Call the classification function
   fits_file_link, predicted_class, probability, image = classifier.single_source(ra,       
                                                                       dec,plot=False)

   # Example for multi-source classification
   input_file = "test.csv"
   output_file = "results.csv"
   classifier.multi_sources(file=input_file, ra_col=0, dec_col=1,
                           output_file=output_file)
   
   

For more information, see the associated research papers:

• MNRAS
• Astro-ph
• GitHub

How to cite:

@article{Alhassan2018,

author = {Alhassan, Wathela and Taylor, A R and Vaccari, Mattia},

doi = {10.1093/mnras/sty2038},

issn = {0035-8711},

journal = {Monthly Notices of the Royal Astronomical Society},

month = {jul},

title = {{The FIRST Classifier: Compact and Extended Radio Galaxy Classification using Deep Convolutional Neural Networks}},

url = {https://academic.oup.com/mnras/advance-article/doi/10.1093/mnras/sty2038/5060783},

year = {2018}

}