Search for the Sunyaev-Zeldovich effect on the maps of microwave background radiation from the Planck space mission.
Authors: O. V. Verkhodanov, A. P. Topchieva, A. D. Oronovskaya, D. A. Shorin, S. A. Bazrov.
The paper proposes a method for searching for objects with the Sunyaev-Zeldovich effect using the convolutional neural network ResNet18. The machine learning method used will make it possible to speed up the selection of candidates for SZ-effect galaxy clusters.
The catalog for identifying radio sources was compiled using the GLESP (Gauss-Legendre Sky Pixelization) pixelation scheme at frequencies: 100, 143, 217, 353, 545 GHz. In this work, it is shown that this approach can be applied to the analysis of data sets, which will make it possible to select the most probable candidates for galaxy clusters at high redshifts using an accelerated procedure.
The authors express their gratitude to the Traektoria Foundation for Support of Scientific, Cultural and Educational Initiatives, V. S. Ivashkin for assistance in carrying out this work.