Data mining techniques on astronomical spectra data – I. Clustering analysis

Yang, Haifeng; Shi, Chenhui; Cai, Jianghui; Zhou, Lichan; Yang, Yuqing; Zhao, Xujun; Yang, He; Hao, Jing

doi:10.1093/mnras/stac2975

Cited by 18 publications

(7 citation statements)

References 102 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Clustering algorithms are unsupervised and should pick up on how certain spectra are grouped together without knowledge of the underlying physical parameters. There are many types of clustering algorithms (see Yang et al 2022 for a recent review of clustering on spectral data); in this work we consider soft, inductive clustering algorithms. A soft algorithm is one that can assign a probability of each training set spectrum belonging to any one of the identified clusters.…”

Section: Clusteringmentioning

confidence: 99%

KilonovAE: Exploring Kilonova Spectral Features with Autoencoders

Ford,

Vieira,

Ruan

et al. 2024

ApJ

View full text Add to dashboard Cite

Kilonovae are likely a key site of heavy r-process element production in the Universe, and their optical/infrared spectra contain insights into both the properties of the ejecta and the conditions of the r-process. However, the event GW170817/AT2017gfo is the only kilonova so far with well-observed spectra. To understand the diversity of absorption features that might be observed in future kilonovae spectra, we use the TARDIS Monte Carlo radiative transfer code to simulate a suite of optical spectra spanning a wide range of kilonova ejecta properties and r-process abundance patterns. To identify the most common and prominent absorption lines, we perform dimensionality reduction using an autoencoder, and we find spectra clusters in the latent space representation using a Bayesian Gaussian Mixture model. Our synthetic kilonovae spectra commonly display strong absorption by strontium 38Sr ii, yttrium 38Y ii, and zirconium 40Zr i–ii, with strong lanthanide contributions at low electron fractions (Y e ≲ 0.25). When a new kilonova is observed, our machine-learning framework will provide context on the dominant absorption lines and key ejecta properties, helping to determine where this event falls within the larger “zoo” of kilonovae spectra.

show abstract

Section: Clusteringmentioning

confidence: 99%

KilonovAE: Exploring Kilonova Spectral Features with Autoencoders

Ford,

Vieira,

Ruan

et al. 2024

ApJ

View full text Add to dashboard Cite

show abstract

“…Deep learning techniques, with their robust representation learning capabilities, have transformed various fields, including astronomy (e.g., Yang et al 2022Yang et al , 2023. The transformer model, introduced by Vaswani et al (2017), stands out with notable success, especially in natural language processing (NLP) and computer vision (CV).…”

Section: Introductionmentioning

confidence: 99%

SPT: Spectral transformer for age and mass estimations of red giant stars

Zhang,

Wu,

et al. 2024

A&A

View full text Add to dashboard Cite

The ages and masses of red giants are key to our understanding of the structure and evolution of the Milky Way. Traditional isochrone methods for these estimations are inherently limited due to overlapping isochrones in the Hertzsprung-Russell diagram, while asteroseismology, albeit more precise, requires high-precision, long-term observations. In response to these challenges, we developed a novel framework, spectral transformer (SPT), to predict the ages and masses of red giants aligned with asteroseismology from their spectra. The main component of SPT is the multi-head Hadamard self-attention mechanism, which is designed specifically for spectra and can capture complex relationships across different wavelengths. Furthermore, we introduced a Mahalanobis distance-based loss function, to address scale imbalance and interaction mode loss, and we incorporated a Monte Carlo dropout for a quantitative analysis of the prediction uncertainty. Trained and tested on 3,880 red giant spectra from LAMOST, the SPT has achieved remarkable age and mass estimations, with average percentage errors of 17.64<!PCT!> and 6.61<!PCT!>, respectively. It has also provided uncertainties for each corresponding prediction. These results significantly outperform traditional machine learning algorithms, demonstrating a high level of consistency with asteroseismology methods and isochrone-fitting techniques. In the future, our work will leverage datasets from the Chinese Space Station Telescope and Large Synoptic Survey Telescope to enhance the precision of the model and broaden its applicability in the fields of astronomy and astrophysics.

show abstract

“…It is also meaningful to explore the "Unknown" spectra to improve the data quality and promote spectral processing techniques by using new data mining methods (Yang et al 2022a(Yang et al , 2023. This paper designs an unsupervised-based analytical framework and performs a detailed analysis of LAMOST "Unknown" spectra.…”

Section: Introductionmentioning

confidence: 99%

An in-depth Exploration of LAMOST Unknown Spectra Based on Density Clustering

Yang

Yin²,

Cai³

et al. 2023

Res. Astron. Astrophys.

View full text Add to dashboard Cite

LAMOST (Large Sky Area Multi-Object Fiber Spectroscopic Telescope) has completed the observation of nearly 20 million celestial objects, including a class of spectra labeled ‘Unknown’. Besides low signal-to-noise ratio, these spectra often show some anomalous features that do not work well with current templates. In this paper, a total of 638,000 ‘Unknown’ spectra from LAMOST DR5 are selected, and an unsupervised-based analytical framework of ‘Unknown’ spectra named SA-Frame (Spectra Analysis-Frame) is provided to explore their origins from different perspectives. The SA-Frame is composed of three parts: NAPC-Spec clustering, characterization and origin analysis. First, NAPCSpec( Nonparametric density clustering algorithm for spectra) characterizes different features in the ”unknown” spectrum by adjusting the influence space and divergence distance to minimize the effects of noise and high dimensionality, resulting in 13 types. Second, characteristic extraction and representation of clustering results are carried out based on spectral lines and continuum, where these 13 types are characterized as regular spectra with low S/Ns, splicing problems, suspected galactic emission signals, contamination from city light and ungregarious type respectively. Third, a preliminary analysis of their origins is made from the characteristics of the observational targets, contamination from the sky, and the working status of the instruments. These results would be valuable for improving the overall data quality of large-scale spectral surveys.

show abstract

Data mining techniques on astronomical spectra data – I. Clustering analysis

Cited by 18 publications

References 102 publications

KilonovAE: Exploring Kilonova Spectral Features with Autoencoders

KilonovAE: Exploring Kilonova Spectral Features with Autoencoders

SPT: Spectral transformer for age and mass estimations of red giant stars

An in-depth Exploration of LAMOST Unknown Spectra Based on Density Clustering

Contact Info

Product

Resources

About