Skip to content

A tutorial on classification and photometric redshift regression of astronomical sources using supervised machine learning techniques.

Notifications You must be signed in to change notification settings

mofokeng-chaka/Classification-Photo-z_Regression_Demo

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

79 Commits
 
 
 
 
 
 

Repository files navigation

Classification and Photo-z Regression of Astronomical Sources

Screenshot 2021-01-10 at 06 58 41 Screenshot 2021-01-10 at 06 51 02

This tutorial uses two machine learning algorithms: random forest and k-nearest neighbour classifiers to classify three different classes (stars, galaxies and quasars) and also estimate photometric redshifts (for galaxies and quasars) by using colour magnitudes as features to learn. The reference dataset uses the corrected photometric magnitudes from SDSS together with the corresponding WISE magnitudes.

The tutorial uses a reference dataset : sdss_all_small_dr16_allwise_red.fits.gz - https://drive.google.com/file/d/1ljutewiejnx_hWujvNY9SzmQm7NnIsXa/view?usp=sharing

Dependencies

Step 1. Download the dataset

Use the link provided above to download the referece dataset sdss_all_small_dr16_allwise_red.fits.gz and place it in the same directory as the jupyter notebook.

Step 2. Run demo

Run all the cells in the notebook to make sure all the dependencies are installed correctly.

Step 3. Improving results

Last step, after completion step 2 is the optimization step. To optimize, tweak the parameters of the models (defined under _classifier and _regressor functions) to produce optimal results

About

A tutorial on classification and photometric redshift regression of astronomical sources using supervised machine learning techniques.

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published