Edit `network_transfer_function` to one best function that matches paper

[axiezai]

• gii paramter in local model
• local/global alpha needs to be distinguished better
• decide on weighted sum for final global version
• tauC (tau_G) needs to be removed
• P(w) * u --> how should we choose the noise?

[axiezai]

I think in the end, we should have 2 network transfer function models, one for the paper, one from HM.

[axiezai]

@parulv1 add on to this list if you see I missed anything

[parulv1]

• Determine the expression for Heid(w).