Skip to content

Detect automatically twin galaxies with machine learning techniques.

Notifications You must be signed in to change notification settings

AIAstronomy/twin-galaxies

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

9 Commits
 
 

Repository files navigation

Twin Galaxies

The aim is to search for galaxies with similar properties. We namely them twin galaxies.

Projects

Python 3.8 Keras 2.2 Jupyter

Author: Carlos García Peral@carlosgp-ai

Co-Author and Advisor: Guzmán-Álvarez, César A. @cguz

Abstract

A novel method to detect similarities between galaxies. We based on different deep learning models to build the most efficient model. The innovative lies in the fact that the vast majority of works are based on the classification and detection of galaxies. Our approach is different, not seeking classification as the final objective but rather the search for galaxies with similar properties. We namely them twin galaxies.

Our approach, first, classifies the galaxies by their morphology, and then based on a Convolutional Neural Network (CNN), compares the feature vectors of the galaxies (same as the works done by Victor and Miguel) and from those vectors, we calculate the Euclidean distance establishing a ranking that will indicate the twin galaxies. We train and test our models using SDSS images for objects in the CALIFA SURVEY.

Python 3.8 Keras 2.2 Jupyter

Author: Víctor Zamora Abarca

Co-Author and Advisor: Guzmán-Álvarez, César A. @cguz

Python 3.8 Keras 2.2 Jupyter

Author: Miguel López Marín

Co-Author and Advisor: Guzmán-Álvarez, César A. @cguz

Publications:

Three Master Thesis degrees:

Working on a conference paper:

Releases

No releases published

Packages

No packages published