9 Hidden Markov Model (20 Points). For a finite-state random sequence {Z1}with the model of {rt, A} and its observation sequence is {X;} , the joint distribution of X and Z with the model 0 is given by p(X, Z|O). a) Summarize the general EM scheme for HMM (E-step and M-step). b) Assuming each observation probability density is Bernoulli, i.e. K P(X,Z10) = 1 [p(z:)P(X|0:)]2 p = i=1 please derive the corresponding model learning procedure under the EM scheme