Estimating Cosmological Parameters and Reconstructing Hubble Constant with Artificial Neural Networks: A Test with Reconstructed $H(z)

In this work, we present a new approach to constrain the cosmological parameters and estimate Hubble constant. We reconstructed a function from observational Hubble data using an Artificial Neural Network (ANN). The training data we used are covariance matrix and mock $H(z)$. With the reconstructed $H(z)$, we can get the Hubble constant, and thus do the comparison with the CMB-based measurements. Furthermore, in order to constrain the cosmological parameters, we sampled data points from the reconstructed data and estimated the posterior distribution. The constraining result behaved well comparing to the ones from the mock observational Hubble data. We propose that the $H(z)$ reconstructed by our artificial neural network can represent the actual distribution of real observational data, and therefore can be used in further cosmological research.