Predicting the Peak Energy of Swift Gamma-Ray Bursts Using Supervised Machine Learning

TL;DR

A method based on the SuperLearner framework that integrates multiple supervised machine learning algorithms to predict the peak energies of Swift/BAT GRBs is proposed, significantly increasing the number of GRBs with known peak energies and providing new statistical support for constraining GRB emission mechanisms and energy origins.

Abstract

Gamma-ray bursts (GRBs) are among the most energetic explosive phenomena in the universe, and their peak energy ($E_{\rm p}$) is a key physical quantity for understanding the prompt emission mechanism. However, due to the limited energy coverage of the Swift satellite, a large fraction of Swift GRBs lack reliable measurements of the peak energy. Therefore, developing an accurate and efficient method to predict $E_{\rm p}$ is of great importance. In this work, we propose a method based on the SuperLearner framework that integrates multiple supervised machine learning algorithms to predict $E_{\rm p}$ of Swift/BAT GRBs. We use the Swift/BAT observational data from December 2004 to September 2022 as training features, and adopt the peak energies of 516 GRBs jointly detected by Swift and either Fermi/GBM or Konus-Wind as training labels. After training and testing multiple supervised models, the final SuperLearner ensemble yields a more robust and reliable predictive model. In 100 iterations of 5-fold cross validation, the predicted $E'_{\rm p}$ values show a tight correlation with the observed $E_{\rm p}$, with an average Pearson correlation coefficient of $r = 0.72$. Compared with previous Bayesian estimates, our model provides predictions that are likely closer to the true values. Based on the trained model, we further predict the peak energies of 650 Swift GRBs, significantly increasing the number of GRBs with known peak energies and providing new statistical support for constraining GRB emission mechanisms and energy origins.

Figures (16)

• Figure 1: Distributions of the four input quantities, $\Gamma$ , $T_{90}$, $S_{\rm \gamma}$, and $F_{\rm p}$, for the training set (blue region) and the generalization set (orange region).
• Figure 2: The distribution of $E_{\rm p}$ in the training set.
• Figure 3: Correlation heatmap of various parameters in the training set.
• Figure 4: Feature importance scores trained by the random forest algorithm, reflecting the relative importance of the four input features in the random forest training set.
• Figure 5: Optimization plot of hyperparameters in the random forest algorithm. Panels (a) and (b) show the variations in RMSE and Pearson correlation coefficient with the number of trees and the maximum tree depth, respectively.