An undercomplete autoencoder with MAE loss denoises SDSS absorption-line spectra and recovers the 4000 Å break, HδF, and Mgb with no large systematic bias, while a Butterworth filter overfits the noise.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
astro-ph.IM 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Emergent Denoising of SDSS Galaxy Spectra Through Unsupervised Deep Learning
An undercomplete autoencoder with MAE loss denoises SDSS absorption-line spectra and recovers the 4000 Å break, HδF, and Mgb with no large systematic bias, while a Butterworth filter overfits the noise.