Autoencoder-based framework for anomaly detection in stellar spectra: application to the MaNGA Stellar Library

TL;DR

This paper introduces an autoencoder-based machine learning framework to detect unusual stellar spectra in the MaNGA Stellar Library, highlighting its effectiveness and limitations in identifying spectral anomalies.

Contribution

It presents a novel application of autoencoders for anomaly detection in stellar spectra, demonstrating its utility on real astronomical data.

Findings

Successfully identified anomalous spectra, including instrumental issues and rare stellar types.

Showed autoencoders can effectively detect spectral anomalies with reconstruction error.

Discussed the limitations and sources of errors in autoencoder-based anomaly detection.

Abstract

A machine-learning-based method is developed to identify objects with unusual stellar spectra. The method employs an autoencoder, a neural network trained to compress spectral data into a low-dimensional representation and subsequently reconstruct it. Spectra that deviate significantly from the dominant patterns in the training dataset are identified using the reconstruction error as an anomaly score. The models are applied to selected datasets from the MaNGA Stellar Library, an empirical library of stellar spectra. Several spectra are flagged as anomalous: an object with likely instrumental and/or reduction issues, two carbon stars, and an oxygen-rich thermally pulsating asymptotic giant branch star. The sources of the large reconstruction errors are examined, and the effectiveness and limitations of autoencoder-based approaches for detecting anomalous stellar spectra are discussed.