KZKe-Rui Zhu

Ke-Rui Zhu

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Ke-Rui Zhu has a total of 18 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Estimating redshift distributions of Fermi blazars with multiple machine-learning models

    by Kerui Zhu, Jiaming Chen, Yonggang Zheng, Ruixin Zhou, Shan Chang, Li Zhang, Niansheng Tang

    Context. Approximately half of the Fermi-LAT-detected blazars lack spectroscopic redshift measurements, which limits population studies and investigations of the cosmological evolution of the gamma-ray blazar population. Aims. We aim to develop a multi-model ensemble framework to systematically evaluate the performance of machine-learning models for blazar redshift estimation and provide predictive redshift probability distributions with quantified uncertainties for all blazars with unknown redshifts. Methods. We constructed an ensemble of 35 machine-learning regression models in the 1/(1 + z ) space. We combined predictions from the individual models to build nonparametric redshift probability density functions (PDFs) for each source, thereby explicitly characterizing the predictive uncertainty. We then stacked the source-level PDFs to derive population-level redshift distributions. Results. We find comparable predictive performance across the diverse models within our framework, indicating a feature-limited performance plateau. We compiled a catalog of predictive redshift distributions for 1590 blazars without measured redshifts. The stacked PDFs recover the characteristic double-peaked redshift structure of blazars without using source-class labels. For blazar candidates of uncertain type, the predicted redshift distribution exhibits a bimodal structure, suggesting a slightly larger contribution from FSRQ-like sources. Conclusions. Our multi-model statistical framework achieves competitive predictive performance while retaining faint sources. We provide predicted redshift distributions with quantified uncertainties for 1590 blazars without measured redshifts, offering improved support for population and evolutionary studies. More broadly, this approach highlights the practical value of uncertainty-aware multi-model regression ensembles in astronomical data analysis.

  • Open Access

    Modeling Gamma-Ray Burst Spectra with Convolutional Neural Networks: Fast-cooling Synchrotron Emission in a Decaying Magnetic Field

    by Jia-Ming Chen, Ke-Rui Zhu, Shan Chang, Zhao-Yang Peng, Yong-Gang Zheng, Li Zhang

    The radiation mechanism of gamma-ray burst (GRB) prompt emission remains uncertain. Although the fast-cooling synchrotron model in a decaying magnetic field can account for the characteristic nonthermal spectral shape, its computational cost has limited its use in systematic observational fitting and statistical model comparison. We develop a convolutional neural network (CNN)-based spectral emulator for this physical model and train it on a large synthetic data set generated over a physically motivated parameter space. The trained network reproduces the numerical spectra with high fidelity while reducing the cost of spectral evaluation to the millisecond level. We then incorporate the emulator into a Bayesian spectral-analysis framework and apply it to the time-resolved spectra of GRB 231020A observed by the Fermi Gamma-ray Burst Monitor. In most time intervals, the decaying-field fast-cooling synchrotron model provides better fits and smaller Bayesian information criterion values than the standard fast-cooling synchrotron model. These results suggest that a radially decaying magnetic field provides a plausible and more physically motivated interpretation of the prompt-emission spectrum of this burst, while also indicating that the emulator offers a practical route for large-sample Bayesian inference and systematic comparisons of GRB prompt-emission models.

  • Open Access

    Fast-cooling Synchrotron in Decaying Magnetic Fields: Implications for the Gamma-Ray Burst Spectral Distribution

    by Jia-Ming Chen, Ke-Rui Zhu, Zhao-Yang Peng, Yong-Gang Zheng, Yun-Lu Gong, Shan Chang, Shi-Ting Tian, Li Zhang

    The prompt emission spectra of gamma-ray bursts (GRBs) are commonly described by the empirical Band function. The typical low-energy spectral index is ∼−1, which poses a challenge to standard synchrotron radiation models. We systematically investigate a fast-cooling synchrotron model with a decaying magnetic field and test, within an observation-consistent pipeline, whether it reproduces the Band-fit parameter distributions in the Gamma-Ray Burst Monitor (GBM) catalog, in a statistical sense. We solve the electron continuity equation with synchrotron, adiabatic, and synchrotron self-Compton (SSC) cooling to obtain the time-dependent electron distribution and synthetic spectra; we then forward-fold through the GBM response matrices and recover ( α , β , E p ) with Band fits. We find that magnetic-field decay can harden the recovered α relative to the fast-cooling limit in part of parameter space, but the effect is not robust and is sensitive to the location of E p within the finite band and to spectral curvature; varying key physical scales reshapes the recovered α distribution, indicating that catalog α often represents an effective in-band slope rather than the asymptotic index. SSC cooling provides modest additional hardening and, in our setups, does not stabilize α near the observed peak. Using Monte Carlo samples designed to mimic the observations, the model yields α mostly between −1.5 and −0.8, but remains centered around α ≈ −1.5. Overall, while a decaying-field fast-cooling synchrotron can partially alleviate overly soft spectra expected from standard fast-cooling synchrotron emission, it still falls short of reproducing the GBM α distribution at the population level, implying that additional physical processes are required.

  • Open Access

    Multiwavelength Variability and Jet Properties of Low-synchrotron-peaked BL Lacertae Objects

    by R. X. Zhou, S. M. Hu, K. R. Zhu, S. J. Kang, Y. G. Zheng

    Low-synchrotron-peaked BL Lac objects (LBLs) are promising candidates for masquerading BL Lacs, and the mechanism responsible for their high-energy γ -ray emission still requires further investigation. In this paper, we conduct detailed analyses of the variability and spectral properties of a sample of LBLs, utilizing 16 yr of multi-instrument observations spanning from 2008 October to 2024 September. To understand the physical properties of LBLs, we model their broadband, nearly simultaneous spectral energy distributions during different variability episodes using a one-zone leptonic scenario. The major outcomes can be outlined as follows. (1) The broadband spectra of LBLs can be reproduced within a leptonic model framework incorporating seed photons from both the broad-line region (BLR) and the dusty torus (DT). The γ -ray emission is inferred to be generated in the location beyond the BLR and within the DT. (2) For an e ± –p jet, the jet power of LBLs follows the relation P e ′ (the kinetic power carried in relativistic electrons) ∼ P p ′ (the kinetic power in cold protons) > P rad ′ (the radiation) > P B ′ ′ (Poynting flux). (3) The LBL radiation mechanism during multiwavelength simultaneous flares can be described by a one-zone leptonic model, while peculiar orphan X-ray flares are more plausibly explained by a stochastic dissipation scenario.


  • Open Access

    The variability properties of extreme blazars: insights from high-synchrotron-peaked BL Lacs

    by R X Zhou, S M Hu, K R Zhu, S J Kang, Y G Zheng

    We have investigated the spectral and variability behaviours of a distinct sample of 16 high-synchrotron-peaked BL Lac objects (HBLs) on long time-scales, utilizing nearly 16 yr of observations from Fermi-LAT, Swift X-Ray Telescope, and Swift-Ultraviolet and Optical Telescope. Based on long-period $\gamma$-ray and X-ray light curves analysis, we processed the quasi-simultaneous multiwavelength observations of the sample in both low- and high-flux states and classified the HBLs according to their spectral index and flux variations. We then adopt a one-zone leptonic model incorporating both isotropic and anisotropic electron distributions to reproduce the quasi-simultaneous spectral energy distributions (SEDs) of the sample across various activity states, thereby probing the underlying radiation mechanisms. Our main results can be summarized as follows. (1) The X-ray and $\gamma$-ray spectra of most HBLs with high variability exhibit an hardening of the spectrum in the high-flux state, accompanied by a shift in the SED peak frequency. (2) Compared to the traditional homogeneous SSC model, the anisotropic SSC model can yield smaller Doppler factors and stronger magnetic fields. (3) HBLs satisfy the relation ${P_{\rm {e}}}$ (the kinetic power carried in relativistic electrons) $\sim$ ${P_{{B^\prime }}}$ (Poynting flux) $\le$ ${P_{{\rm {rad}}}}$ (the radiative power) $<$ ${P_{\rm {p}}}$ (the kinetic power in cold protons) in different activity states in the anisotropic one-zone SSC model.


  • Open Access

    GRB 240619A: Evidence for a Compact Star Merger Origin

    by Jia-Ming Chen, Ke-Rui Zhu, Xue-Zhao Chang, Zhao-Yang Peng, Li Zhang

    GRB 240619A, detected by Fermi-Gamma-ray Burst Monitor with a duration of ∼36.1 s at redshift z = 0.3965, a short, hard main emission (ME), a quiescent interval, and a soft extended emission (EE). We perform a detailed analysis of the temporal and spectral properties of both phases using Bayesian inference and machine learning techniques. Our main results show: (1) the spectral lags are 0.27 ± 0.03 s (ME) and 0.13 ± 0.19 s (EE), with the ME phase following the lag–luminosity relation; (2) The minimum variability timescales are ∼0.04 s (ME) and ∼0.48 s (EE), consistent with short gamma-ray bursts (SGRB) and long gamma-ray bursts (LGRB) characteristics, respectively, suggesting that the radiation in the two phases may originate from different physical mechanisms. (3) the effective amplitude f eff ≈ 3.18 suggests an intrinsic origin consistent with SGRBs; (4) the spectral hardness, peak energy, Amati/Yonetoku relations, and energy–hardness–duration classification indicate that the ME phase is more similar to SGRBs, while the EE phase resembles LGRBs; (5) the local event rate density ( 4.6 6 − 3.86 + 11.25 × 1 0 − 3 Gpc −3 yr −1 ) further highlights its peculiarity. Spectral analysis indicates that its nonthermal emission can be well explained by the fast-cooling synchrotron radiation model. Curvature effect analysis suggests jet acceleration in the EE phase, with the jet becoming Poynting-flux dominated. The combined evidence suggests that GRB 240619A is an SGRB with EE, likely originating from a compact binary merger, and exhibiting physical properties similar to those of GRB 211211A.


  • Open Access

    Classification and Physical Characteristic Analysis of Fermi-GBM Gamma-Ray Bursts Based on Deep Learning

    by Jia-Ming Chen, Ke-Rui Zhu, Zhao-Yang Peng, Li Zhang

    The classification of gamma-ray bursts (GRBs) has long been an unresolved problem. Early long- and short-burst classification based on duration is not convincing owing to the significant overlap in duration plot, which leads to different views on the classification results. We propose a new classification method based on convolutional neural networks and adopt a sample including 3774 GRBs observed by Fermi-GBM to address the T 90 overlap problem. By using count maps that incorporate both temporal and spectral features as inputs, we successfully classify 593 overlapping events into two distinct categories, thereby refuting the existence of an intermediate GRB class. Additionally, we apply the optimal model to extract features from the count maps and visualize the extracted GRB features using the t-SNE algorithm, discovering two distinct clusters corresponding to S-type and L-type GRBs. To further investigate the physical properties of these two types of bursts, we conduct a time-integrated spectral analysis and discover significant differences in their spectral characteristics. The analysis also shows that most GRBs associated with kilonovae belong to the S type, while those associated with supernovae are predominantly L type, with few exceptions. Additionally, the duration characteristics of short bursts with extended emission suggest that they may manifest as either L-type or S-type GRBs. Compared to traditional classification methods (Amati and energy–hardness–duration methods), the new approach demonstrates significant advantages in classification accuracy and robustness without relying on redshift observations. The deep learning classification strategy proposed in this paper provides a more reliable tool for future GRB research.


  • Open Access

    Classification and physical characteristics analysis of Fermi-GBM Gamma-ray bursts based on Deep-learning

    by Jia-Ming Chen, Ke-Rui Zhu, Zhao-Yang Peng, Li Zhang

    The classification of Gamma-Ray Bursts has long been an unresolved problem. Early long and short burst classification based on duration is not convincing due to the significant overlap in duration plot, which leads to different views on the classification results. We propose a new classification method based on Convolutional Neural Networks and adopt a sample including 3774 GRBs observed by Fermi-GBM to address the $T_\text{90}$ overlap problem. By using count maps that incorporate both temporal and spectral features as inputs, we successfully classify 593 overlapping events into two distinct categories, thereby refuting the existence of an intermediate GRB class. Additionally, we apply the optimal model to extract features from the count maps and visualized the extracted GRB features using the t-SNE algorithm, discovering two distinct clusters corresponding to S-type and L-type GRBs. To further investigate the physical properties of these two types of bursts, we conduct a time-integrated spectral analysis and discovered significant differences in their spectral characteristics. The analysis also show that most GRBs associated with kilonovae belong to the S-type, while those associated with supernovae are predominantly L-type, with few exceptions. Additionally, the duration characteristics of short bursts with extended emission suggest that they may manifest as either L-type or S-type GRBs. Compared to traditional classification methods (Amati and EHD methods), the new approach demonstrates significant advantages in classification accuracy and robustness without relying on redshift observations. The deep learning classification strategy proposed in this paper provides a more reliable tool for future GRB research.

  • Open Access

    The Statistical Analysis of GeV Spectral Breaks in Bright Gamma-Ray Flat-spectrum Radio Quasars

    by K. R. Zhu, J. M. Chen, L. Zhang

    We present the statistical results of GeV spectral breaks of bright gamma-ray flat-spectrum radio quasars (FSRQs) in the energy range of 0.1–10 GeV based on New Pass 8 data of the Large Area Telescope on board Fermi Gamma-Ray Space Telescope. We have fitted the 15 yr average gamma-ray spectra of 755 FSRQs by using both the broken power law (BPL) and logarithmic parabola (LP) models, and obtained 87 bright gamma-ray FSRQs with their integrated photon fluxes greater than 2.16 × 10−8 cm−2 s−1. From our results, the FSRQ population shows similar preferences for both the BPL and LP models in gamma-ray spectral fitting, and the clustering analysis suggests that BPL-preferred and LP-preferred FSRQs belong to the same category. Our results indicate that GeV spectral breaks in bright gamma-ray FSRQs are located at 2.90 ± 1.92 GeV in the rest frame, and the observed change in photon index is Δγ = 0.45 ± 0.19, which is consistent with the expected value for a cooling break of electrons scattering seed photons.


  • Open Access

    GRB 231129C: Another Thermal Emission Dominated Gamma-Ray Burst

    by Jia-Ming Chen, Ke-Rui Zhu, Zhao-Yang Peng, Li Zhang

    This study presents detailed time-integrated and time-resolved spectral analysis of the Fermi Gamma-ray Burst Monitor observations of the bright GRB 231129C. The results reveal its distinct spectral characteristics, featuring a hard low-energy spectral index (α) and soft high-energy spectral index (β), similar to GRB 090902B, suggesting a possible dominance of thermal emission. Further analysis indicates that 92% of the spectral indices exceed the synchrotron “line of death,” with the hardest index at α ∼ +0.44. Simultaneously, 53% of the spectra can be well fitted by the nondissipative photosphere model, supporting a potential origin from a nondissipative photosphere. Additionally, we observe strong correlations between the spectral index α and peak energy E p with flux. For the α−F relationship, we employ F = F 0 e (3.00±0.10)α to describe it, whereas the E p−F relationship requires a smoothly bending power-law function. Based on the framework proposed by Hascoët et al. and Gao & Zhang, the jet characteristics of this burst were studied, revealing that both methods support the suitability of a pure fireball model for this GRB at small initial jet radii.