This set of Notebooks are designed to create a pipeline of Machine Learning models for predicting the detection of Radio Galaxies and their redshift values.
One intial model is trained to classify between galaxies and AGN. A second model is trained to classify between AGN having, or not, radio detection above certain limit (given by detections in selected radio surveys). A third model is trained to predict redshift values for radio-detected AGN.
The pipeline works by applying the first model to obtain a list of predicted AGN. Then, the second model is applied to these predicted AGN and radio-detected sources are predicted. Finally, the third model is applied to the predicted radio-detected AGN and a predicted redshift value is obtained.
Datasets are located in the relative path (not available in this repository):
../../Catalogs/
Description of most files and folders can be seen in files_naming.txt.
Plots and images are not included in this repository.