Multidimensional Data Driven Classification of Emission-line Galaxies

TL;DR

This paper introduces a novel multidimensional soft clustering method using Gaussian mixtures and SVMs to classify emission-line galaxies based on four optical-line ratios, enhancing accuracy and probabilistic classification.

Contribution

The paper presents a new multidimensional clustering approach combining Gaussian mixture models and SVMs for galaxy classification, utilizing all four emission-line ratios simultaneously.

Findings

High classification accuracy achieved with the proposed method.

Soft classification provides probabilistic membership for galaxies.

Method captures complex multi-dimensional spectral structures.

Abstract

We propose a new soft clustering scheme for classifying galaxies in different activity classes using simultaneously 4 emission-line ratios; log([NII ]/Ha), log([SII]/Ha), log([OI]/Ha) and log([OIII]/Hb). We fit 20 multivariate Gaussian distributions to the 4-dimensional distribution of these lines obtained from the Sloan Digital Sky Survey (SDSS) in order to capture local structures and subsequently group the multivariate Gaussian distributions to represent the complex multi-dimensional structure of the joint distribution of galaxy spectra in the 4 dimensional line ratio space. The main advantages of this method are the use of all four optical-line ratios simultaneously and the adoption of a clustering scheme. This maximises the available information, avoids contradicting classifications, and treats each class as a distribution resulting in soft classification boundaries and providing the probability for an object to belong to each class. We also introduce linear multi-dimensional decision surfaces using support vector machines based on the classification of our soft clustering scheme. This linear multi-dimensional hard clustering technique shows high classification accuracy with respect to our soft-clustering scheme.