Forward–backward algorithm: Difference between revisions

Revision as of 13:45, 5 April 2013

Revision as of 13:55, 5 April 2013

Line 304:

observations = ('normal', 'cold', 'dizzy')

transition_probability = {

start_probability = {'Healthy': 0.6, 'Fever': 0.4}

transition_probability = {

'Healthy' : {'Healthy': 0.69, 'Fever': 0.3, 'E': 0.01},

'Fever' : {'Healthy': 0.4, 'Fever': 0.59, 'E': 0.01},

}

start_probability = {'Healthy': 0.6, 'Fever': 0.4}

emission_probability = {

'Healthy' : {'normal': 0.5, 'cold': 0.4, 'dizzy': 0.1},

Line 323 ⟶ 321:

L = len(x)

fwd = []

f_prev = {}

# Run forward

for il, x_i in enumerate(x):

f_curr fwd = {}[]

f_prev = {}

f_curr = {}

for st in states:

if i == 0:

# Initialize base fwd cases

prev_f_sum = a_0[st]

fwd.append(f_curr)

f_prev = f_curr

p_fwd = sum(f_curr[kl]*a[kl][end_st] for k in states)

bkw = []

Line 342 ⟶ 339:

# Run bkw

p_bkw = sum(a_0[l1] * e[l][x[01]] * b_curr[l0] for l in states)

posterior = {}

Line 365 ⟶ 362:

def example():

return fwd_bkw(observations,

states,)

start_probability,

transition_probability,

emission_probability,

end_state)

for line in example():

print ' '.join(map(str, line))