When we estimate the parameters of the multinomial logit model. we use what is called maximum likelihood estimation. Often times, people maximize log-likelihood instead of the likelihood itself. Which of the following is NOT a correct statement regarding this method? 0 Log-likelihood maximization yields the same optimal parameters as likelihood maximization. In other words: arg maxp log (I: (p)) = arg maxp (11 (p)) O The reason log-likelihood maximization and likelihood maximization yield the same optimal parameter is because logarithm is a monotonic increasing function. O From a computing complexity point of View, handling log-likelihood is often easier than handling likelihood since likelihood could become very small. O In the case of the multinomial logit model, we use log-likelihood instead of likelihood to get an expression which is always linear in x