EMアルゴリズムの例
8281 ワード
#coding:utf-8
import math
import copy
import numpy as np
import matplotlib.pyplot as plt
isdebug = True
# k , k=2。
# 2 Sigma, Mu1,Mu2。
# 1000
# , 6,40,20,2
# 6,
# 20, 40
# 1000
def ini_data(Sigma,Mu1,Mu2,k,N):
#
global X
#
global Mu
#
global Expectations
#1*N , N
X = np.zeros((1,N))
# ,
# ,
Mu = np.random.random(2) #0-1
print Mu
# 1000*2 ,
Expectations = np.zeros((N,k))
# N
for i in xrange(0,N):
# 0.5 1 , 0.5 2
if np.random.random(1) > 0.5:
# 40 , N(40,Sigma)
X[0,i] = np.random.normal()*Sigma + Mu1 #
else:
# 40 , N(20,Sigma)
X[0,i] = np.random.normal()*Sigma + Mu2
if isdebug:
print "***********"
print u" X:"
print X
#E
# : Sigma, k, N
def e_step(Sigma,k,N):
#
global Expectations
#
global Mu
#
global X
# ,
for i in xrange(0,N):
# ,
Denom = 0
# ,
for j in xrange(0,k):
#
Denom += math.exp((-1/(2*(float(Sigma**2))))*(float(X[0,i]-Mu[j]))**2)
# ,
for j in xrange(0,k):
#
Numer = math.exp((-1/(2*(float(Sigma**2))))*(float(X[0,i]-Mu[j]))**2)
#
Expectations[i,j] = Numer/Denom
if isdebug:
print "***********"
print u" E(Z):"
print len(Expectations)
#
print Expectations.size
#
print Expectations.shape
#
print Expectations
#M
def m_step(k,N):
# P(k|xi)
global Expectations
#
global X
#
#
for j in xrange(0,k):
Numer = 0
Denom = 0
# ,
#
for i in xrange(0,N):
# P(k|xi)xi
Numer += Expectations[i,j]*X[0,i]
# Nk,Nk P(k|xi)
Denom +=Expectations[i,j]
# uk
Mu[j] = Numer / Denom
# iter_num , Epsilon
# 1000 , 0.0001
# : Sigma, Mu1, Mu2
# k, N, iter_num, Epsilon
def run(Sigma,Mu1,Mu2,k,N,iter_num,Epsilon):
#
ini_data(Sigma,Mu1,Mu2,k,N)
print u" <u1,u2>:", Mu
# 1000
for i in range(iter_num):
#
Old_Mu = copy.deepcopy(Mu)
#E
e_step(Sigma,k,N)
#M
m_step(k,N)
#
print i,Mu
#
if sum(abs(Mu-Old_Mu)) < Epsilon:
break
if __name__ == '__main__':
#sigma,mu1,mu2, , , ,
run(6,40,20,2,1000,1000,0.0001)
plt.hist(X[0,:],100) #
plt.show()