(a) Consider the LDA model X|Y = k ~ N(uk, 02), where k = 1 or 0. For simplicity, we assume * is univariate. Prove that the LDA model can be rewritten as a logistic regression. To show this result, you may want to compute what is the log odds under the LDA model. Recall that the log odds is defined as log[P(Y = 1|X = x) / P(Y = 0|X = x)]. (b) Similarly, consider the QDA model X|Y = k ~ N(uk, o ), where k = 1 or 0. For simpli- city, we assume X is univariate. Prove that the QDA model can be rewritten as a logistic regression with some nonlinear terms