In search of an interaction in the dark sector through Gaussian Process and ANN approaches (2505.04336v1)

Abstract: Whether the current observational data indicate any evidence of interaction between the dark sector is a matter of supreme interest at the present moment. This article searched for an interaction in the dark sector between a pressure-less dark matter and a dark energy fluid with constant equation of state, $w_{\rm DE}$. For this purpose, two non-parametric approaches, namely, the Gaussian Process (GP) and the Artificial Neural Networks (ANN) have been employed and using the Hubble data from Cosmic Chronometers (CC), Pantheon+ from Supernovae Type Ia and their combination we have reconstructed the interaction function. We find that for $w_{\rm DE} =-1$, the interaction in the dark sector is not prominent while for $w_{\rm DE} \neq -1$, evidence of interaction is found depending on the value of $w_{\rm DE}$. In particularly, we find that if we start deviating from $w_{\rm DE} = -1$ either in the quintessence ($w_{\rm DE} > -1$) or phantom ($w_{\rm DE} < -1$) direction, an emergence of dark interaction is observed from both GP and ANN reconstructions. We further note that ANN which is applied for the first time in this context seems to play a very efficient role compared to GP.