隐马尔可夫模型python实现

最新推荐文章于 2024-05-09 01:18:03 发布

captain飞虎大队

最新推荐文章于 2024-05-09 01:18:03 发布

阅读量1.6k

点赞数

分类专栏： python 人工神经网络

本文链接：https://blog.csdn.net/weixin_41529093/article/details/110631765

版权

python 同时被 2 个专栏收录

453 篇文章 18 订阅

订阅专栏

人工神经网络

5 篇文章 1 订阅

订阅专栏

仅作为保存，大佬请跳过。

文章目录

前向算法
后向算法
参考

直接上代码

前向算法

from numpy import *
import numpy as np
import matplotlib as plt
import math


def hmm_forward():
    A=np.array([[0.5,0.1,0.4],[0.3,0.5,0.2],[0.2,0.2,0.6]])
    B=np.array([[0.5,0.5],[0.4,0.6],[0.7,0.3]])
    pai=np.array([0.2,0.3,0.5])
    o=np.array([0,1,0,0,1,0,1,1])

    # 初始化
    T=len(o)
    boxnumber=shape(B)[0]
    alpha=np.zeros(T*boxnumber).reshape(boxnumber,T)

    for i in range(boxnumber):
        alpha[i][0]=pai[i]*B[i][o[0]]


    # 迭代
    sigma=0
    for j in range(T-1):
        for i in range(boxnumber):
            for k in range(boxnumber):
                sigma+=alpha[k][j]*A[k][i]
            alpha[i][j+1]=sigma*B[i][o[j+1]]
            sigma=0


    # 终止
    prob=np.sum(alpha,axis=0)

    print(alpha)
    print(prob[-1])


if __name__=='__main__':
    hmm_forward()

后向算法

from numpy import *
import numpy as np
import matplotlib as plt
import math


def hmm_backward():
    A=np.array([[0.5,0.1,0.4],[0.3,0.5,0.2],[0.2,0.2,0.6]])
    B=np.array([[0.5,0.5],[0.4,0.6],[0.7,0.3]])
    pai=np.array([0.2,0.3,0.5])
    o=np.array([0,1,0,0,1,0,1,1])

    # 初始化
    T=len(o)
    boxnumber=shape(B)[0]
    beta=np.zeros(T*boxnumber).reshape(boxnumber,T)

    for i in range(boxnumber):
        beta[i][0]=1


    # 迭代
    sigma=0
    t=T
    for j in range(T-1):
        t -= 1
        for i in range(boxnumber):
            for k in range(boxnumber):
                sigma+=A[i][k]*B[k][o[t]]*beta[k][j]
            beta[i][j+1]=sigma
            sigma=0

    print(beta)


    # 终止
    sigma=0
    for i in range(boxnumber):
        sigma+=pai[i]*B[i][o[0]]*beta[i][T-1]
    print(sigma)



if __name__=='__main__':
    hmm_backward()