Definition

\(Y\) is distributed \(Bernoulli(p)\) and

\[g^{-1}(\eta) = \frac{1}{1 + e^{-\eta}}\] That is,

\[r(x)=E[Y|X=x]=p(x)=\frac{1}{1 + e^{-\beta_0 - \beta_1x}}\] given that the \[E[Y] = p\]

for \(Bernoulli(p)\).

So, we can also write

\[g(E[Y|X=x]) = log \frac{p}{1-p}= \beta_0 + \beta_1x\]