SVR Algorithm as a Tool for More Optimal Intergalactic Medium Simulation in the Epoch of Reionization

S. Mobina Hosseini, Mahsa Berahman, Seyed Sajad Tabasi, Javad T. Firouzjaee

Published 2023-10-26Version 1

All kinds of simulations of the intergalactic medium, such as hydrodynamic simulation, N-body simulation, numerical and semi-numerical simulation, etc., have been used to realize the history of this medium. One of these simulations is 21SSD, which is specifically focused on the epoch of reionization. This simulation deepens our understanding of the physics behind the intergalactic medium by considering the free parameters related to the Wouthuysen-Field coupling fluctuations and X-ray and Lyman line transfers in the intergalactic medium, and by presenting the plots of the power spectrum, brightness temperature, etc. in different redshifts. However, due to many physical phenomena that play significant roles in this epoch, simulations of the intergalactic medium are usually extremely complex, time-consuming, and require very powerful hardware. In this work, by using the Support Vector Regression algorithm and based on the 21SSD simulation datasets, we have tried to make the machine fully understand the brightness temperature changes in terms of redshift for different astrophysical free parameters values. At first, we trained the machine with the results of the 21SSD simulation. Then, the machine was able to predict the brightness temperature in terms of redshift with very high accuracy for other interval coefficients. Although we have used this algorithm to estimate the brightness temperature, it seems that this algorithm can be easily used for other parts of cosmology and astrophysics. With its help, it is possible to save time and obtain results with extraordinary accuracy similar to complex simulations, even with normal hardware.