隐马尔科夫模型 使用（下）

# !/usr/bin/python
# -*- coding:utf-8 -*-

import numpy as np
from hmmlearn import hmm
import matplotlib.pyplot as plt
import matplotlib as mpl
from sklearn.metrics.pairwise import pairwise_distances_argmin
import warnings


def expand(a, b): d = (b - a) * 0.05 return a-d, b+d


if __name__ == "__main__": warnings.filterwarnings("ignore")   # hmmlearn(0.2.0) < sklearn(0.18) np.random.seed(0) n = 5   # 隐状态数目 n_samples = 1000 pi = np.random.rand(n) pi /= pi.sum() print('初始概率：', pi) A = np.random.rand(n, n) mask = np.zeros((n, n), dtype=np.bool) mask[0][1] = mask[0][4] = True mask[1][0] = mask[1][2] = True mask[2][1] = mask[2][3] = True mask[3][2] = mask[3][4] = True mask[4][0] = mask[4][3] = True A[mask] = 0 for i in range(n): A[i] /= A[i].sum() print('转移概率：
  

 

  
1
  
2
  
3
  
4
  
5
  
6
  
7
  
8
  
9
  
10
  
11
  
12
  
13
  
14
  
15
  
16
  
17
  
18
  
19
  
20
  
21
  
22
  
23
  
24
  
25
  
26
  
27
  
28
  
29
  
30
  
31
  
32
  
33
  
34
  
35
  
36
 

文章来源: maoli.blog.csdn.net，作者：刘润森！，版权归原作者所有，如需转载，请联系作者。

原文链接：maoli.blog.csdn.net/article/details/89219503