Spectral Synthesis via Mean Field approach Independent Component Analysis

TL;DR

This paper introduces a Mean Field Bayesian ICA method for galaxy spectral analysis, enabling efficient spectral decomposition and accurate physical parameter recovery, even for low signal-to-noise data.

Contribution

The paper presents a novel MF-ICA algorithm for galaxy spectral analysis that improves efficiency and accuracy over existing methods.

Findings

MF-ICA compresses spectral libraries into few independent components.

The method accurately recovers galaxy physical parameters.

It performs well on low signal-to-noise spectra from DEEP2 data.

Abstract

In this paper, we apply a new statistical analysis technique, Mean Field approach to Bayesian Independent Component Analysis (MF-ICA), on galaxy spectral analysis. This algorithm can compress the stellar spectral library into a few Independent Components (ICs), and galaxy spectrum can be reconstructed by these ICs. Comparing to other algorithms which decompose a galaxy spectrum into a combination of several simple stellar populations, MF-ICA approach offers a large improvement in the efficiency. To check the reliability of this spectral analysis method, three different methods are used: (1) parameter-recover for simulated galaxies, (2) comparison with parameters estimated by other methods, and (3) consistency test of parameters from the Sloan Digital Sky Survey galaxies. We find that our MF-ICA method not only can fit the observed galaxy spectra efficiently, but also can recover the physical parameters of galaxies accurately. We also apply our spectral analysis method to the DEEP2 spectroscopic data, and find it can provide excellent fitting for those low signal-to-noise spectra.