4. Logistic regression: In a logistic model, our observed data is a random variable Y that takes values y : 0 or y = 1 and depends on some set of explanatory variables [3'0 + Zj jxj. The probability of observing y = 1, conditioned on the independent variables 1:, is equal to ]E[Y|X] = 1/(1+exp((Bo +23- 33-13). Derive the link function for this model, i.e., the function g() such that g(lE[Y|X]) : 50 + Zj jIj. (Hint: Don't let the notation distract you. Let u 2 g + :5; Bjmj and w : lE[Y[X] and solve in : 1/(1 + exp(u)) for n. You can identify 9 by n = g(w).)