Yihan Tao
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Yihan Tao has a total of 93 co-authors. Here are the top co-authors by number of shared publications:
Chenzhou Cui
Chinese Academy of Sciences
6 shared publications- Chenzhou Cui
National Astronomical Observatories, Chinese Academy of Sciences
5 shared publications Ali Luo
Chinese Academy of Sciences
3 shared publicationsChanghua Li
National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China;
3 shared publicationsDongwei Fan
National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China;
3 shared publications- Xiao Kong
Chinese Academy of Sciences
3 shared publications
Latest Publications
- 5 October 2026
- Published by Cambridge University Press (CUP) in Proceedings of the International Astronomical Union
- Vol. 21 (S397) , 169-173
FALCO: towards Foundation model of Astronomical Light Curves for time dOmain astronomy
Time domain astronomy confronts challenges in the rapid identification and characterisation of diverse transients and variables within large-scale surveys, demanding efficient computational methods beyond conventional task-specific approaches. We propose FALCO - Foundation model of Astronomical Light Curves for time dOmain astronomy - to address the challenges by establishing generalisable representations of light curve data through self-supervised learning. Initial implementation of the model employs a modified GPT-2 architecture trained on Kepler light curves via a prediction task. Evaluations across stellar surface gravity inference, variable source classification, and stellar flare identification demonstrate that leveraging learned representations from FALCO achieves competitive or superior performance with reduced training samples compared to task-specific baselines. Future developments will extend the model to more complex scenarios, including multi-survey integration and irregular temporal sampling patterns.
- 5 October 2026
- Published by Cambridge University Press (CUP) in Proceedings of the International Astronomical Union
- Vol. 21 (S397) , 169-173
FALCO: towards Foundation model of Astronomical Light Curves for time dOmain astronomy
Time domain astronomy confronts challenges in the rapid identification and characterisation of diverse transients and variables within large-scale surveys, demanding efficient computational methods beyond conventional task-specific approaches. We propose FALCO - Foundation model of Astronomical Light Curves for time dOmain astronomy - to address the challenges by establishing generalisable representations of light curve data through self-supervised learning. Initial implementation of the model employs a modified GPT-2 architecture trained on Kepler light curves via a prediction task. Evaluations across stellar surface gravity inference, variable source classification, and stellar flare identification demonstrate that leveraging learned representations from FALCO achieves competitive or superior performance with reduced training samples compared to task-specific baselines. Future developments will extend the model to more complex scenarios, including multi-survey integration and irregular temporal sampling patterns.
- Open Access
- 28 July 2026
- Published by American Astronomical Society in The Astrophysical Journal Supplement Series
- Vol. 285 (2) , 55
An AI-ready Stellar Dataset Assembled from LAMOST DR11
We present a clean, homogeneous, and AI-ready stellar dataset assembled from LAMOST DR11 low-resolution spectra, tailored for data-driven models. This dataset contains 111,417 high-quality spectra for 91,542 AFGK-type stars, with stellar labels such as T eff , log g , [Fe/H], [ α /M], stellar mass, and absolute G -band magnitude. It reaches a metallicity of [Fe/H] ∼ −3.8 dex, significantly below the LAMOST pipeline’s lower limit of −2.5 dex, and covers T eff ≈ 3800–13,000 K, log g ≈ 0.1–5.0 dex, [ α /M] from −0.4 to 0.6 dex, mass from 0.4 to 6.0 M ⊙ , and M G from −4.0 to 8.0 mag. Self-consistency validation using neural networks confirms the reliability of stellar labels and shows only minor differences between the adopted and network-predicted parameters. By combining this dataset with the 5105 M-type stars from our previous work, we trained a neural network model and applied it to the complete set of LAMOST DR12 AFGKM-type spectra, obtaining stellar parameters for 9 million spectra. The model yields precise estimates that agree well with APOGEE and Gaia estimates, with uncertainties of about 55 K in T eff , 0.10 dex in log g , 0.07 dex in [Fe/H], 0.02 dex in [ α /M], a relative error of about 4.2% in stellar mass, and 0.13 mag in M G . Distances inferred from M G show a relative error of 7.4%. This work delivers a public, AI-ready dataset and a value-added catalog for LAMOST DR12, extends investigations of very metal-poor stars toward lower metallicities, and supports direct, efficient estimation of stellar parameters, mass, and distance from large spectroscopic surveys.
- 28 July 2026
- Published by American Astronomical Society in The Astrophysical Journal Supplement Series
- Vol. 285 (2) , 55
Open AccessAn AI-ready Stellar Dataset Assembled from LAMOST DR11
We present a clean, homogeneous, and AI-ready stellar dataset assembled from LAMOST DR11 low-resolution spectra, tailored for data-driven models. This dataset contains 111,417 high-quality spectra for 91,542 AFGK-type stars, with stellar labels such as T eff , log g , [Fe/H], [ α /M], stellar mass, and absolute G -band magnitude. It reaches a metallicity of [Fe/H] ∼ −3.8 dex, significantly below the LAMOST pipeline’s lower limit of −2.5 dex, and covers T eff ≈ 3800–13,000 K, log g ≈ 0.1–5.0 dex, [ α /M] from −0.4 to 0.6 dex, mass from 0.4 to 6.0 M ⊙ , and M G from −4.0 to 8.0 mag. Self-consistency validation using neural networks confirms the reliability of stellar labels and shows only minor differences between the adopted and network-predicted parameters. By combining this dataset with the 5105 M-type stars from our previous work, we trained a neural network model and applied it to the complete set of LAMOST DR12 AFGKM-type spectra, obtaining stellar parameters for 9 million spectra. The model yields precise estimates that agree well with APOGEE and Gaia estimates, with uncertainties of about 55 K in T eff , 0.10 dex in log g , 0.07 dex in [Fe/H], 0.02 dex in [ α /M], a relative error of about 4.2% in stellar mass, and 0.13 mag in M G . Distances inferred from M G show a relative error of 7.4%. This work delivers a public, AI-ready dataset and a value-added catalog for LAMOST DR12, extends investigations of very metal-poor stars toward lower metallicities, and supports direct, efficient estimation of stellar parameters, mass, and distance from large spectroscopic surveys.
- Open Access
- 3 December 2025
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 171 (1)
FALCO: Foundation Model of Astronomical Light Curves for Time Domain Astronomy. Implementation and Applications on Kepler Data
Time-domain surveys have advanced astronomical research by revealing diverse variable phenomena, from stellar flares to transient events. The scale and complexity of survey data, along with the demand for rapid classification, present significant challenges for analysis. While machine learning offers solutions, most existing models are tailored to single tasks, struggle to generalize, and depend heavily on large, accurately labeled datasets. This paper presents an initial implementation of FALCO, a foundation model trained on Kepler light curves via self-supervised learning using a Transformer-based architecture. The model has been evaluated on Kepler data across three distinct light-curve analysis tasks, and it demonstrates robust performance in all tasks, achieving an accuracy of 95% for stellar variability classification across eight classes, an overall RMSE of 0.1305 dex in surface gravity estimation (with significantly improved precision of RMSE < 0.08 dex at the low-gravity end where log g < 1, and 0.02 dex near log g ≈ 3), and a precision of 87% in flare identification. These results highlight the versatility of the foundation model in extracting generalizable representations from light curves, enabling easy adaptation to diverse tasks and making it a promising tool for time-domain analysis. Further analysis of model scaling and input light-curve sequence length reveals that larger models and longer input sequences improve performance. We have also applied the model to produce a comprehensive catalog of surface gravity (log g ) measurements for 179,732 Kepler stellar targets, using their light curves.
- 3 December 2025
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 171 (1)
Open AccessFALCO: Foundation Model of Astronomical Light Curves for Time Domain Astronomy. Implementation and Applications on Kepler Data
Time-domain surveys have advanced astronomical research by revealing diverse variable phenomena, from stellar flares to transient events. The scale and complexity of survey data, along with the demand for rapid classification, present significant challenges for analysis. While machine learning offers solutions, most existing models are tailored to single tasks, struggle to generalize, and depend heavily on large, accurately labeled datasets. This paper presents an initial implementation of FALCO, a foundation model trained on Kepler light curves via self-supervised learning using a Transformer-based architecture. The model has been evaluated on Kepler data across three distinct light-curve analysis tasks, and it demonstrates robust performance in all tasks, achieving an accuracy of 95% for stellar variability classification across eight classes, an overall RMSE of 0.1305 dex in surface gravity estimation (with significantly improved precision of RMSE < 0.08 dex at the low-gravity end where log g < 1, and 0.02 dex near log g ≈ 3), and a precision of 87% in flare identification. These results highlight the versatility of the foundation model in extracting generalizable representations from light curves, enabling easy adaptation to diverse tasks and making it a promising tool for time-domain analysis. Further analysis of model scaling and input light-curve sequence length reveals that larger models and longer input sequences improve performance. We have also applied the model to produce a comprehensive catalog of surface gravity (log g ) measurements for 179,732 Kepler stellar targets, using their light curves.
- 16 October 2025
- Published by Springer Nature
- p. 51-67
MAVI: MLLM-Enhanced Anomaly Validator and Interpreter for Astronomical Time Series
- 16 October 2025
- Published by Springer Nature
- p. 51-67
MAVI: MLLM-Enhanced Anomaly Validator and Interpreter for Astronomical Time Series
- Open Access
- 1 September 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 11, 9501-9509
FLARE: A Framework for Stellar Flare Forecasting Using Stellar Physical Properties and Historical Records
Stellar flare events are critical observational samples for astronomical research; however, recorded flare events remain limited. Stellar flare forecasting can provide additional flare event samples to support research efforts. Despite this potential, no specialized models for stellar flare forecasting have been proposed to date. In this paper, we present extensive experimental evidence demonstrating that both stellar physical properties and historical flare records are valuable inputs for flare forecasting tasks. We then introduce FLARE (Forecasting Light-curve-based Astronomical Records via features Ensemble), the first-of-its-kind large model specifically designed for stellar flare forecasting. FLARE integrates stellar physical properties and historical flare records through a novel Soft Prompt Module and Residual Record Fusion Module. Experiments on the Kepler light curve dataset demonstrate that FLARE achieves superior performance compared to other methods across all evaluation metrics. Finally, we validate the forecast capability of our model through a comprehensive case study.
- 1 September 2025
- Published by International Joint Conferences on Artificial Intelligence
- Vol. 11, 9501-9509
Open AccessFLARE: A Framework for Stellar Flare Forecasting Using Stellar Physical Properties and Historical Records
Stellar flare events are critical observational samples for astronomical research; however, recorded flare events remain limited. Stellar flare forecasting can provide additional flare event samples to support research efforts. Despite this potential, no specialized models for stellar flare forecasting have been proposed to date. In this paper, we present extensive experimental evidence demonstrating that both stellar physical properties and historical flare records are valuable inputs for flare forecasting tasks. We then introduce FLARE (Forecasting Light-curve-based Astronomical Records via features Ensemble), the first-of-its-kind large model specifically designed for stellar flare forecasting. FLARE integrates stellar physical properties and historical flare records through a novel Soft Prompt Module and Residual Record Fusion Module. Experiments on the Kepler light curve dataset demonstrate that FLARE achieves superior performance compared to other methods across all evaluation metrics. Finally, we validate the forecast capability of our model through a comprehensive case study.
- 30 May 2025
- Published by IOP Publishing in Research in Astronomy and Astrophysics
- Vol. 25 (6) , 065011
A Retrieval-augmented Analysis Framework for Target Galaxy Search
The identification of specific galaxy populations in large-scale spectroscopic surveys represents an essential yet challenging task, particularly for rare or anomalous galaxies that deviate from the typical galaxy distributions. Traditional methods based on template-fitting or predefining spectral features face challenges in addressing the complexity and scale of modern astronomical data sets. To overcome these limitations, we propose GalSpecEncoder-KB, a modular and flexible framework that combines deep learning with knowledge base retrieval to enable efficient and interpretable analysis of galaxy spectra. The framework integrates a Transformer-based feature encoder, GalSpecEncoder, pre-trained with masked-modeling strategy to capture semantically rich and context-aware spectral representations. By leveraging a Retrieval-Augmented Analysis approach, the knowledge base constructed from catalogs enables metadata retrieval and weighted voting for target galaxy identification. Using the Sloan Digital Sky Survey as a comprehensive case study, we demonstrate the capabilities of the framework for target galaxy search. Experimental results demonstrate the exceptional generalizability and adaptability across diverse galaxy search tasks, including identification of LINERs, Strong Gravitational Lenses, and detection of Outliers, while maintaining robust performance and interpretable spectral analysis capabilities.
- 30 May 2025
- Published by IOP Publishing in Research in Astronomy and Astrophysics
- Vol. 25 (6) , 065011
A Retrieval-augmented Analysis Framework for Target Galaxy Search
The identification of specific galaxy populations in large-scale spectroscopic surveys represents an essential yet challenging task, particularly for rare or anomalous galaxies that deviate from the typical galaxy distributions. Traditional methods based on template-fitting or predefining spectral features face challenges in addressing the complexity and scale of modern astronomical data sets. To overcome these limitations, we propose GalSpecEncoder-KB, a modular and flexible framework that combines deep learning with knowledge base retrieval to enable efficient and interpretable analysis of galaxy spectra. The framework integrates a Transformer-based feature encoder, GalSpecEncoder, pre-trained with masked-modeling strategy to capture semantically rich and context-aware spectral representations. By leveraging a Retrieval-Augmented Analysis approach, the knowledge base constructed from catalogs enables metadata retrieval and weighted voting for target galaxy identification. Using the Sloan Digital Sky Survey as a comprehensive case study, we demonstrate the capabilities of the framework for target galaxy search. Experimental results demonstrate the exceptional generalizability and adaptability across diverse galaxy search tasks, including identification of LINERs, Strong Gravitational Lenses, and detection of Outliers, while maintaining robust performance and interpretable spectral analysis capabilities.
- Open Access
- 5 May 2025
- Published by Springer Nature in Research Square
Galaxy Circus: A New Paradigm for Anomalous Galaxy Discovery with Artificial Intelligence and Citizen Science
Modern telescopes yield unprecedented volumes of observational data, creating a significant challenge for efficiently identifying anomalous astronomical objects or events that could extend the boundaries of human knowledge. Hence, we propose a novel research paradigm that integrates artificial intelligence with citizen science to address the challenge. Specifically, a Bayesian neural network with a pre-trained Astronomical Large Vision Model (ALVM) first pre-screens data for anomalous galaxy candidates, which are then forwarded to citizen scientists for initial identifications in the manner of a “Pair Up” game. At last, with assistance from an astronomical natural language model (ANLM) with retrieval augmented generation (RAG), experts examine the targets selected by citizen scientists and send the results of the analysis back to the citizen scientists for further learning. In our initial experimental application, named Galaxy Circus, 77 qualified citizen scientists have analysed 414391 galaxy images from the DESI Legacy Imaging Surveys with the help of ALVM in the "Pair Up" game, successfully identifying 1377 anomalous galaxies, including 387 (28.10%) merging galaxy candidates, 318 (23.09%) low-surface-brightness galaxy candidates, 3 (0.22%) gravitational lensing candidates, and 669 (48.58%) objects that defy conventional galaxy classification. Notably, 293 galaxies within the 669 galaxies have been previously documented in academic literature as abnormal galaxies. Through further vetting by 6 experts, our approach has identified 1084 new anomalous galaxies with over 9000 times faster than classical anomalous galaxy discovery paradigm (reducing the time required from approximately 75560 hours to just 8.8 hours), which requires scientists to vet each galaxy image, acceleration in scientific research, demonstrating that AI-citizen science integration provides a novel, efficient and reliable paradigm for identifying rare astronomical phenomena in the big data era. Galaxy circus: https://nadc.china-vo.org/galaxycircus/dist1-en
- 5 May 2025
- Published by Springer Nature in Research Square
Open AccessGalaxy Circus: A New Paradigm for Anomalous Galaxy Discovery with Artificial Intelligence and Citizen Science
Modern telescopes yield unprecedented volumes of observational data, creating a significant challenge for efficiently identifying anomalous astronomical objects or events that could extend the boundaries of human knowledge. Hence, we propose a novel research paradigm that integrates artificial intelligence with citizen science to address the challenge. Specifically, a Bayesian neural network with a pre-trained Astronomical Large Vision Model (ALVM) first pre-screens data for anomalous galaxy candidates, which are then forwarded to citizen scientists for initial identifications in the manner of a “Pair Up” game. At last, with assistance from an astronomical natural language model (ANLM) with retrieval augmented generation (RAG), experts examine the targets selected by citizen scientists and send the results of the analysis back to the citizen scientists for further learning. In our initial experimental application, named Galaxy Circus, 77 qualified citizen scientists have analysed 414391 galaxy images from the DESI Legacy Imaging Surveys with the help of ALVM in the "Pair Up" game, successfully identifying 1377 anomalous galaxies, including 387 (28.10%) merging galaxy candidates, 318 (23.09%) low-surface-brightness galaxy candidates, 3 (0.22%) gravitational lensing candidates, and 669 (48.58%) objects that defy conventional galaxy classification. Notably, 293 galaxies within the 669 galaxies have been previously documented in academic literature as abnormal galaxies. Through further vetting by 6 experts, our approach has identified 1084 new anomalous galaxies with over 9000 times faster than classical anomalous galaxy discovery paradigm (reducing the time required from approximately 75560 hours to just 8.8 hours), which requires scientists to vet each galaxy image, acceleration in scientific research, demonstrating that AI-citizen science integration provides a novel, efficient and reliable paradigm for identifying rare astronomical phenomena in the big data era. Galaxy circus: https://nadc.china-vo.org/galaxycircus/dist1-en
- Open Access
- 1 November 2024
- Published by American Astronomical Society in The Astrophysical Journal Supplement Series
- Vol. 275 (1)
LEAVES: An Expandable Light-curve Data Set for Automatic Classification of Variable Stars
With the increasing amount of astronomical observation data, it is an inevitable trend to use artificial intelligence methods for automatic analysis and identification of light curves for full samples. However, data sets covering all known classes of variable stars that meet all research needs are not yet available. There is still a lack of standard training data sets specifically designed for any type of light-curve classification, but existing light-curve training sets or data sets cannot be directly merged into a large collection. Based on the open data sets of the All-Sky Automated Survey for SuperNovae, Gaia, and Zwicky Transient Facility, we construct a compatible light-curve data set named LEAVES for automated recognition of variable stars, which can be used for training and testing new classification algorithms. The data set contains a total of 977,953 variable and 134,592 nonvariable light curves, in which the supported variables are divided into six superclasses and nine subclasses. We validate the compatibility of the data set through experiments and employ it to train a hierarchical random forest classifier, which achieves a weighted average F1-score of 0.95 for seven-class classification and 0.93 for 10-class classification. Experimental results prove that the classifier is more compatible than the classifier established based on a single band and a single survey, and has wider applicability while ensuring classification accuracy, which means it can be directly applied to different data types with only a relatively small loss in performance compared to a dedicated model.
- 1 November 2024
- Published by American Astronomical Society in The Astrophysical Journal Supplement Series
- Vol. 275 (1)
Open AccessLEAVES: An Expandable Light-curve Data Set for Automatic Classification of Variable Stars
With the increasing amount of astronomical observation data, it is an inevitable trend to use artificial intelligence methods for automatic analysis and identification of light curves for full samples. However, data sets covering all known classes of variable stars that meet all research needs are not yet available. There is still a lack of standard training data sets specifically designed for any type of light-curve classification, but existing light-curve training sets or data sets cannot be directly merged into a large collection. Based on the open data sets of the All-Sky Automated Survey for SuperNovae, Gaia, and Zwicky Transient Facility, we construct a compatible light-curve data set named LEAVES for automated recognition of variable stars, which can be used for training and testing new classification algorithms. The data set contains a total of 977,953 variable and 134,592 nonvariable light curves, in which the supported variables are divided into six superclasses and nine subclasses. We validate the compatibility of the data set through experiments and employ it to train a hierarchical random forest classifier, which achieves a weighted average F1-score of 0.95 for seven-class classification and 0.93 for 10-class classification. Experimental results prove that the classifier is more compatible than the classifier established based on a single band and a single survey, and has wider applicability while ensuring classification accuracy, which means it can be directly applied to different data types with only a relatively small loss in performance compared to a dedicated model.
- Open Access
- 1 November 2024
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 168 (6) , 233
A Photometric Redshift Catalogue of Galaxies from the DESI Legacy Imaging Surveys DR10
The large-scale measurement of galaxy redshifts holds significant importance for cosmological research, as well as for understanding the formation and evolution of galaxies. This study utilizes a known sample obtained by cross-correlating the Dark Energy Spectroscopic Instrument (DESI) Legacy Imaging Surveys DR10 galaxy catalog with various galaxy catalogs from different spectroscopic surveys. The methods Easy and Accurate Photometric Redshifts from Yale (EAZY) and CatBoost are employed to estimate redshifts. In the case of EAZY, the known sample is used solely for testing, while CatBoost utilizes it for both training and testing purposes. The known sample is categorized into different subsamples based on various classification methods. Several CatBoost regression models are trained and optimized using these subsamples. By comparing the performance of different methods and models, it is observed that the two-step and two-part models outperform the one-step model, with further enhancements achieved through the combination of the two-step and two-part models. Based on the findings from all experiments, we propose a photometric redshift estimation workflow designed to facilitate the photometric redshift measurement of all galaxies within the DESI Legacy Imaging Surveys DR10. Consequently, a photometric redshift catalog has been released, comprising a total of 1,533,107,988 galaxies. Among these, 312,960,837 galaxies have reliable redshift estimates, determined using the CatBoost algorithm, with magnitude limits set at g > 24.0, r > 23.4, and z > 22.5. For galaxies with g, r, and z magnitudes exceeding these thresholds, the photometric redshifts estimated by EAZY can be employed as a reference.
- 1 November 2024
- Published by American Astronomical Society in The Astronomical Journal
- Vol. 168 (6) , 233
Open AccessA Photometric Redshift Catalogue of Galaxies from the DESI Legacy Imaging Surveys DR10
The large-scale measurement of galaxy redshifts holds significant importance for cosmological research, as well as for understanding the formation and evolution of galaxies. This study utilizes a known sample obtained by cross-correlating the Dark Energy Spectroscopic Instrument (DESI) Legacy Imaging Surveys DR10 galaxy catalog with various galaxy catalogs from different spectroscopic surveys. The methods Easy and Accurate Photometric Redshifts from Yale (EAZY) and CatBoost are employed to estimate redshifts. In the case of EAZY, the known sample is used solely for testing, while CatBoost utilizes it for both training and testing purposes. The known sample is categorized into different subsamples based on various classification methods. Several CatBoost regression models are trained and optimized using these subsamples. By comparing the performance of different methods and models, it is observed that the two-step and two-part models outperform the one-step model, with further enhancements achieved through the combination of the two-step and two-part models. Based on the findings from all experiments, we propose a photometric redshift estimation workflow designed to facilitate the photometric redshift measurement of all galaxies within the DESI Legacy Imaging Surveys DR10. Consequently, a photometric redshift catalog has been released, comprising a total of 1,533,107,988 galaxies. Among these, 312,960,837 galaxies have reliable redshift estimates, determined using the CatBoost algorithm, with magnitude limits set at g > 24.0, r > 23.4, and z > 22.5. For galaxies with g, r, and z magnitudes exceeding these thresholds, the photometric redshifts estimated by EAZY can be employed as a reference.
- 1 August 2024
- Published by IOP Publishing in Research in Astronomy and Astrophysics
- Vol. 24 (8) , 085016
X-Ray Source Classification Using Machine Learning: A Study with EP-WXT Pathfinder LEIA
X-ray observations play a crucial role in time-domain astronomy. The Einstein Probe (EP), a recently launched X-ray astronomical satellite, emerges as a forefront player in the field of time-domain astronomy and high-energy astrophysics. With a focus on systematic surveys in the soft X-ray band, EP aims to discover high-energy transients and monitor variable sources in the universe. To achieve these objectives, a quick and reliable classification of observed sources is essential. In this study, we developed a machine learning classifier for autonomous source classification using data from the EP-WXT Pathfinder—Lobster Eye Imager for Astronomy (LEIA) and EP-WXT simulations. The proposed Random Forest classifier, built on selected features derived from light curves, energy spectra, and location information, achieves an accuracy of approximately 95% on EP simulation data and 98% on LEIA observational data. The classifier is integrated into the LEIA data processing pipeline, serving as a tool for manual validation and rapid classification during observations. This paper presents an efficient method for the classification of X-ray sources based on single observations, along with implications of most effective features for the task. This work facilitates rapid source classification for the EP mission and also provides valuable insights into feature selection and classification techniques for enhancing the efficiency and accuracy of X-ray source classification that can be adapted to other X-ray telescope data.
- 1 August 2024
- Published by IOP Publishing in Research in Astronomy and Astrophysics
- Vol. 24 (8) , 085016
X-Ray Source Classification Using Machine Learning: A Study with EP-WXT Pathfinder LEIA
X-ray observations play a crucial role in time-domain astronomy. The Einstein Probe (EP), a recently launched X-ray astronomical satellite, emerges as a forefront player in the field of time-domain astronomy and high-energy astrophysics. With a focus on systematic surveys in the soft X-ray band, EP aims to discover high-energy transients and monitor variable sources in the universe. To achieve these objectives, a quick and reliable classification of observed sources is essential. In this study, we developed a machine learning classifier for autonomous source classification using data from the EP-WXT Pathfinder—Lobster Eye Imager for Astronomy (LEIA) and EP-WXT simulations. The proposed Random Forest classifier, built on selected features derived from light curves, energy spectra, and location information, achieves an accuracy of approximately 95% on EP simulation data and 98% on LEIA observational data. The classifier is integrated into the LEIA data processing pipeline, serving as a tool for manual validation and rapid classification during observations. This paper presents an efficient method for the classification of X-ray sources based on single observations, along with implications of most effective features for the task. This work facilitates rapid source classification for the EP mission and also provides valuable insights into feature selection and classification techniques for enhancing the efficiency and accuracy of X-ray source classification that can be adapted to other X-ray telescope data.