Deep Learning application for stellar parameters determination: III- Denoising Procedure
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866917858939764736 |
|---|---|
| author | Gebran, Marwan Bentley, Ian Brienza, Rose Paletou, Frédéric |
| author_facet | Gebran, Marwan Bentley, Ian Brienza, Rose Paletou, Frédéric |
| contents | In this third paper in a series, we investigate the need of spectra denoising for the derivation of stellar parameters. We have used two distinct datasets for this work. The first one contains spectra in the range of 4450-5400 Å at a resolution of 42000 and the second in the range of 8400-8800 Å at a resolution of 11500. We constructed two denoising techniques, an autoencoder, and a Principal Component Analysis. Using random Gaussian noise added to synthetic spectra, we have trained a Neural Network to derive the stellar parameters Teff, log g, ve sin i, ξt, and [M/H] of the denoised spectra. We find that, independently of the denoising technique, the stellar parameters accuracy values do not improve once we denoise the synthetic spectra. This is true with and without applying data augmentation to the stellar parameters Neural Network. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2412_04631 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Deep Learning application for stellar parameters determination: III- Denoising Procedure Gebran, Marwan Bentley, Ian Brienza, Rose Paletou, Frédéric Solar and Stellar Astrophysics Astrophysics of Galaxies Instrumentation and Methods for Astrophysics Computational Physics In this third paper in a series, we investigate the need of spectra denoising for the derivation of stellar parameters. We have used two distinct datasets for this work. The first one contains spectra in the range of 4450-5400 Å at a resolution of 42000 and the second in the range of 8400-8800 Å at a resolution of 11500. We constructed two denoising techniques, an autoencoder, and a Principal Component Analysis. Using random Gaussian noise added to synthetic spectra, we have trained a Neural Network to derive the stellar parameters Teff, log g, ve sin i, ξt, and [M/H] of the denoised spectra. We find that, independently of the denoising technique, the stellar parameters accuracy values do not improve once we denoise the synthetic spectra. This is true with and without applying data augmentation to the stellar parameters Neural Network. |
| title | Deep Learning application for stellar parameters determination: III- Denoising Procedure |
| topic | Solar and Stellar Astrophysics Astrophysics of Galaxies Instrumentation and Methods for Astrophysics Computational Physics |
| url | https://arxiv.org/abs/2412.04631 |