Hidden Markov Model
Consider the HMM on the right.
"Rainy" and "Sunny" are hidden states.
Start
"Walk", "Shop", "Clean" are observations.
0.4
0.6
a. Using the Forward Algorithm, compute
02
Rainy
Sunny 0.6
the probability of the observation sequence
0.4
[Clean, Shop, Walk].
0.
0.6
0.5
0.1
0.
b. Using the Viterbi Algorithm, find the most likely sequence of hidden states if we are given the observation sequence
Walk
Clean
Shop
[Clean, Shop, Walk].
For both (a) and (b) above, draw their trellis, and show the alphas (Forward Algorithm) and v's (Viterbi Algorithm), similar to the examples on the slides.