Abstract: Core-collapse supernovae (CCSNe) are one of the most anticipated future gravitational wave (GW) sources. The GW emission is dominated by the oscillation modes of the newly born proto-neutron star (PNS) and the stalled accretion shock. The stochastic nature of such a signal makes its interpretation a very challenging task and template matching techniques are inadequate. Asteroseismology can serve as an alternative way to infer properties of the PNS from GW observations. The oscillations can be described by a system of partial differential equations, which can be solved as an eigenvalue problem. I am going to present two independent methods of solving them, a machine learning approach based on physics informed neural networks (PINNs) and spectral methods. In addition, my analysis includes, for the first time, both an accretion flow and a standing accretion shock. In doing so, I add some of the missing ingredients towards a more realistic PNS asteroseismology.