Model-independent reconstruction of f (<i>T</i> ) gravity using genetic algorithms*

Abstract In this paper, we use genetic algorithms, a specific machine learning technique, to achieve a model-independent reconstruction of gravity. By using data derived from cosmic chronometers and the radial Baryon Acoustic Oscillation method, including the latest Dark Energy Spectroscopic Instrument (DESI) data, we reconstruct the Hubble rate, which is the basis parameter for reconstructing gravity without any assumptions. In this reconstruction process, we use the current value of the Hubble rate, , derived by genetic algorithms. The reconstructed function is consistent with the standard ΛCDM cosmology within the 1 confidence level across a broad temporal range. The mean curve, adopting a quadratic form, prompts us to parametrize it using a second-degree polynomial. This quadratic deviation from the ΛCDM scenario is mildly favored by the data.