Discriminative Models $(6$ points$)$: Recall that Logistic Regression is a Discriminative Model.$$In other words, given a set of parameters $\backslash$theta $,$ we model p$($y$|$x$,\backslash$theta $).$ Consider a slightly different binary$$classification model where:$$p$($y$|$x$,\backslash$theta $)$ exp$(($y $\backslash$theta T x$)2/\backslash$sigma $2)(1)$Here x$,\backslash$theta in Rd$,$ y in $\{1,+1\}.$ Answer the following questions:$$ Derive the decision boundaries in the above case.$$ Derive the Conditional MLE estimator of $\backslash$theta $.$ What would be a good choice for the prior model $($Hint: what would be the conjugate prior in$$this case?$).$ Derive the MAP estimator with this prior.$$ Finally, is this a good choice for classification? IF Yes, why and is No$,$ why not?