python實現簡單神經網絡算法，供大家參考，具體內容如下

python實現二層神經網絡

包括輸入層和輸出層

import numpy as np  #sigmoid function def nonlin(x, deriv = False):   if(deriv == True):     return x*(1-x)   return 1/(1+np.exp(-x))  #input dataset x = np.array([[0,0,1],        [0,1,1],        [1,0,1],        [1,1,1]])  #output dataset y = np.array([[0,0,1,1]]).T  np.random.seed(1)  #init weight value syn0 = 2*np.random.random((3,1))-1  for iter in xrange(100000):   l0 = x             #the first layer,and the input layer    l1 = nonlin(np.dot(l0,syn0))  #the second layer,and the output layer     l1_error = y-l1    l1_delta = l1_error*nonlin(l1,True)    syn0 += np.dot(l0.T, l1_delta) print "outout after Training:" print l1 

import numpy as np  #sigmoid function def nonlin(x, deriv = False):   if(deriv == True):     return x*(1-x)   return 1/(1+np.exp(-x))  #input dataset x = np.array([[0,0,1],        [0,1,1],        [1,0,1],        [1,1,1]])  #output dataset y = np.array([[0,0,1,1]]).T  np.random.seed(1)  #init weight value syn0 = 2*np.random.random((3,1))-1  for iter in xrange(100000):   l0 = x             #the first layer,and the input layer    l1 = nonlin(np.dot(l0,syn0))  #the second layer,and the output layer     l1_error = y-l1    l1_delta = l1_error*nonlin(l1,True)    syn0 += np.dot(l0.T, l1_delta) print "outout after Training:" print l1 

這里，
l0:輸入層

l1:輸出層

syn0:初始權值

l1_error:誤差

l1_delta:誤差校正系數

func nonlin:sigmoid函數

可見迭代次數越多，預測結果越接近理想值，當時耗時也越長。

python實現三層神經網絡

包括輸入層、隱含層和輸出層

import numpy as np  def nonlin(x, deriv = False):   if(deriv == True):     return x*(1-x)   else:     return 1/(1+np.exp(-x))  #input dataset X = np.array([[0,0,1],        [0,1,1],        [1,0,1],        [1,1,1]])  #output dataset y = np.array([[0,1,1,0]]).T  syn0 = 2*np.random.random((3,4)) - 1 #the first-hidden layer weight value syn1 = 2*np.random.random((4,1)) - 1 #the hidden-output layer weight value  for j in range(60000):   l0 = X            #the first layer,and the input layer    l1 = nonlin(np.dot(l0,syn0)) #the second layer,and the hidden layer   l2 = nonlin(np.dot(l1,syn1)) #the third layer,and the output layer     l2_error = y-l2    #the hidden-output layer error    if(j%10000) == 0:     print "Error:"+str(np.mean(l2_error))    l2_delta = l2_error*nonlin(l2,deriv = True)    l1_error = l2_delta.dot(syn1.T)   #the first-hidden layer error    l1_delta = l1_error*nonlin(l1,deriv = True)    syn1 += l1.T.dot(l2_delta)   syn0 += l0.T.dot(l1_delta) print "outout after Training:" print l2 

import numpy as np  def nonlin(x, deriv = False):   if(deriv == True):     return x*(1-x)   else:     return 1/(1+np.exp(-x))  #input dataset X = np.array([[0,0,1],        [0,1,1],        [1,0,1],        [1,1,1]])  #output dataset y = np.array([[0,1,1,0]]).T  syn0 = 2*np.random.random((3,4)) - 1 #the first-hidden layer weight value syn1 = 2*np.random.random((4,1)) - 1 #the hidden-output layer weight value  for j in range(60000):   l0 = X            #the first layer,and the input layer    l1 = nonlin(np.dot(l0,syn0)) #the second layer,and the hidden layer   l2 = nonlin(np.dot(l1,syn1)) #the third layer,and the output layer     l2_error = y-l2    #the hidden-output layer error    if(j%10000) == 0:     print "Error:"+str(np.mean(l2_error))    l2_delta = l2_error*nonlin(l2,deriv = True)    l1_error = l2_delta.dot(syn1.T)   #the first-hidden layer error    l1_delta = l1_error*nonlin(l1,deriv = True)    syn1 += l1.T.dot(l2_delta)   syn0 += l0.T.dot(l1_delta) print "outout after Training:" print l2