李航统计学习方法 隐马尔科夫 用前向算法实现例10.2

import numpy as np

# 状态转移矩阵
A = np.array([[0.5,0.2,0.3],[0.3,0.5,0.2],[0.2,0.3,0.5]])
# 观测矩阵
B = np.array([[0.5,0.5],[0.4,0.6],[0.7,0.3]])
# 初始矩阵
pi = np.array([0.2,0.4,0.4])
# 序列
O = np.array([0,1,0])



def forward(A,B,pi,O):
    m=np.shape(A)[0]
    # n=np.shape(B)[1]
    T=len(O)
    alpha=np.zeros((T,m))

    # 计算初值
    for i in range(m):
        alpha[0][i]=pi[i]*B[i][O[0]]
    # 递推
    for i in range(1,T):
        for j in range(m):
            sum=0
            for k in range(m):
                sum+=alpha[i-1][k]*A[k][j]
            alpha[i][j]=sum*B[j][O[i]]
    return alpha

# 统计最后的概率
def cal(alpha):
    c=len(alpha)
    sum=0
    for i in range(c):
        sum+=alpha[-1][i]
    return sum

alpha_i=forward(A,B,pi,O)
P=cal(alpha_i)
print(P)

前向算法理论部分在李航统计学习方法中有较为仔细的介绍，实践部分依据伪代码尽行实现