Title: Non-parametric modeling of the cosmological data, base on the $χ^2$ distribution

Abstract: In the $\Lambda$CDM model, cosmological observations from the late and recent universe reveal a puzzling $\sim 4.5\sigma$ tension in the current rate of universe expansion. In addition to the various scenarios suggested to resolve the tension, non-parametric modeling may provide useful insights. In this paper, we look at three well-known non-parametric methods, the smoothing method, the genetic algorithm, and the Gaussian process. Considering these three methods, we employ the recent Hubble parameters data to reconstruct the rate of universe expansion and supernovae Pantheon sample to reconstruct the luminosity distance. In contrast to the similar studies in the literature, the chi-squared distribution has been used to construct a reliable criterion to select a reconstruction. Finally, we compute the current rate of universe expansion ($H_0$) for each method, provide some discussions regarding the performance of each approach, and compare the results.