2. Derive an EM algorithm to find the maximum likelihood estimate of 0 where 0 is a parameter in the multinomial distribution: (x1, x2, x3, x4) ~ M(n; 0.250, 0.25(2 + 0), 0.5(1 20), 0.50) Similar to the case covered in the class, choose a variable for the missing data and derive the EM algorithm for iteratively estimating 0. Implement this algorithm in R and test it on the dataset (x1, x2, x3, x4) = (6, 52, 28, 14) and n = 100. (hint: partition 0.25(2+0) to 0.5 and 0.250).