Viterbi算法Python实现版

 维比特算法实际是用动态规划解马尔可夫模型的预测问题，即用动态规划求最大概率问题·，这时一条路径对于着一个状态序列。

算法如下： 

def viterbi(i,n,t):
    if t==0:
        return p[i]*b[i,o[0]],''
    max_d=[]
    for j in range(n):
        max_d.append(viterbi(j,n,t-1)[0]*a[j,i])
    m=max_d.index(max(max_d))
    return max(max_d)*b[i,o[t]],str(viterbi(m,n,t-1)[1])+str(m+1)

测试代码：

a=np.array(
    [
        [0.5,0.2,0.3],
        [0.3,0.5,0.2],
        [0.2,0.3,0.5]
    ]
)
b=np.array(
    [
        [0.5,0.5],
        [0.4,0.6],
        [0.7,0.3]
    ]
)
o=[0,1,0,1]
p=[0.2,0.4,0.4]

print(viterbi(0,3,3))
print(viterbi(1,3,3))
print(viterbi(2,3,3))

输出结果：

(0.0018899999999999998, '321')
(0.0030239999999999993, '322')
(0.0022049999999999995, '333')