Identification of 4FGL Uncertain Sources at Higher Resolutions with Inverse Discrete Wavelet Transform

Abstract

Abstract In the forthcoming era of big astronomical data, it is a burden to find target sources from ground-based and space-based telescopes. Although machine-learning methods have been extensively utilized to address this issue, the incorporation of in-depth data analysis can significantly enhance the efficiency of identifying target sources when dealing with massive volumes of astronomical data. In this work, we focused on the task of finding active galactic nucleus (AGN) candidates and identifying BL Lacertae objects (BL Lac) or flat spectrum radio quasar (FSRQ) candidates from the 4FGL_DR3 uncertain sources. We studied the correlations among the attributes of the 4FGL_DR3 catalog and proposed a novel method, named fractal dimension–inverse discrete wavelet transform (FDIDWT), to transform the original data. The transformed data set is characterized as low-dimensional and feature-highlighted, with the estimation of correlation features by fractal dimension theory and the multi-resolution analysis by inverse discrete wavelet transform (IDWT). Combining the FDIDWT method with an improved lightweight MatchboxConv1D model, we accomplished two missions: (1) to distinguish the AGNs from others (non-AGNs) in the 4FGL_DR3 uncertain sources with an accuracy of 96.65% ± 1.32%, namely Mission A; and (2) to classify blazar candidates of uncertain type into BL Lacs or FSRQs with an accuracy of 92.03% ± 2.2%, namely Mission B. There are 1354 AGN candidates in Mission A, and 482 BL Lacs candidates and 128 FSRQ candidates were found in Mission B. The results show a high consistency of greater than 98% with the results in previous works. In addition, our method has the advantage of finding less variable and relatively faint sources than ordinary methods.