隐马尔可夫模型预测算法——维特比算法的Python实现

import numpy as np
def viterbi(A, B, pi, O, Q, V, T):
    delta = np.zeros(len(Q))
    F = np.zeros((len(Q), T))
    for j in range(len(Q)):
        delta[j] = pi[j] * B[j, V[O[0]]]
    for t in range(1, T):
        for i in range(len(Q)):
            delta[i] = np.max(delta * A[:, i]) * B[i, V[O[t]]]
            F[i, t] = np.argmax(delta * A[:, i])
    P = np.max(delta)
    I_T = np.argmax(delta)
    I = [int(i) for i in F[I_T]]
    I.pop(0)
    I.append(I_T)
    return P, I_T, I

Q = [1, 2, 3]
V = {'红':0, '白':1}
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]])
pi = [0.2, 0.4, 0.4]
T = 3
O = ['红', '白', '红']
print(viterbi(A, B, pi, O, Q, V, T))
# (0.014699999999999998, 2, [2, 2, 2])，因为我设置的时间T是从0到2，总共T=3，状态序列也是从0到2，所以最终的最优状态路径索引为2，其实是最后一个状态3