?

?

python實現神經網絡模型算法

?

?

今天，厾羅和大家分享用Python實現神經網絡模型算法，僅用于技術學習交流。

?

實現技巧

?

1.導入依賴庫

主要是安裝相關的依賴庫。本文實現的環境為：python 3.7。

from __future__ import division    
import math      
import random    
import pandas as pd  

2.構建BP神經網絡類

主要是構建三層反向傳播神經網絡類。

""" 三層反向傳播神經網絡 """
class NN:def __init__(self, ni, nh, no):self.ni = ni + 1                            # 輸入層節點self.nh = nh + 1                    # 隱藏層節點self.no = no                      # 輸出層種類self.ai = [1.0] * self.ni    self.ah = [1.0] * self.nh    self.ao = [1.0] * self.no    self.wi = self.makeMatrix(self.ni, self.nh)  # 輸出層到隱藏層的映射矩陣self.wo = self.makeMatrix(self.nh, self.no)  # 隱藏層到輸出層的映射矩陣for i in range(self.ni):          for j in range(self.nh):    self.wi[i][j] = self.rand(-0.2, 0.2)  for j in range(self.nh):for k in range(self.no):self.wo[j][k] = self.rand(-2, 2)  #前向傳播，激活神經網絡的所有節點def update(self, inputs):if len(inputs) != self.ni - 1:print(len(inputs),self.ni - 1)raise ValueError('與輸入層節點數不符！')    for i in range(self.ni - 1):    self.ai[i] = inputs[i]    for j in range(self.nh):                  # self.nh表示隱藏層的節點數sum = 0.0                            # 激活項a = g(z)  z = Θ^T x ;sum相當于z，每次循環歸零for i in range(self.ni):                  #通過循環z = Θ^T x ，因為Θ、x均為向量sum = sum + self.ai[i] * self.wi[i][j]  #〖 Z〗^((2))=Θ^((1)) a^((1))self.ah[j] = self.sigmoid(sum)    # a^((2))=g(z^((2)))，這里使用sigmoid()函數作為激活函數for k in range(self.no):sum = 0.0for j in range(self.nh):sum = sum + self.ah[j] * self.wo[j][k]  #〖 Z〗^((3))=Θ^((2)) a^((2))self.ao[k] = self.sigmoid(sum)    # a^((3))=g(z^((3)))return self.ao[:]#反向傳播,計算節點激活項的誤差def backPropagate(self, targets, lr):               # targets為某樣本實際種類分類，lr為梯度下降算法的學習率output_deltas = [0.0] * self.nofor k in range(self.no):error = targets[k] - np.round_(self.ao[k])output_deltas[k] = self.dsigmoid(self.ao[k]) * error# 計算隱藏層的誤差hidden_deltas = [0.0] * self.nh    for j in range(self.nh):error = 0.0for k in range(self.no):error = error + output_deltas[k] * self.wo[j][k]    hidden_deltas[j] = self.dsigmoid(self.ah[j]) * error# 更新輸出層權重for j in range(self.nh):            # 反向傳播算法，求出每個節點的誤差后，反向更新權重for k in range(self.no):change = output_deltas[k] * self.ah[j]    self.wo[j][k] = self.wo[j][k] + lr * change   # 更新輸入層權重for i in range(self.ni):                    for j in range(self.nh):change = hidden_deltas[j] * self.ai[i]self.wi[i][j] = self.wi[i][j] + lr * change# 計算誤差error = 0.0for k in range(self.no):                                    error += 0.5 * (targets[k] - np.round_(self.ao[k])) ** 2  return error                                          #用測試集輸出準確率def test(self, patterns):                            count = 0num=0for p in patterns:target = p[1]result = self.update(p[0])                    print(p[0], ':', target, '->', np.round_(result))num=0for k in range(self.no):if (target[k] == np.round_(result[k])):num +=1print(num)if num==3:count +=1print("******************",(target) == (np.round_(result)),"******************")accuracy = int(float(count / len(patterns))*100)print('accuracy: %-.9f' % accuracy,"%")      #輸出訓練過后神經網絡的權重矩陣def weights(self):print('輸入層權重:')for i in range(self.ni):print(self.wi[i])print()print('輸出層權重:')for j in range(self.nh):print(self.wo[j])#用訓練集訓練神經網絡def train(self, patterns, iterations=1000, lr=0.1):  for i in range(iterations):error = 0.0                    for p in patterns:            inputs = p[0]            targets = p[1]            self.update(inputs)          error = error + self.backPropagate(targets, lr)  if i % 100 == 0:print("percent：",int(i/iterations*100),"%",'   error: %-.9f' % error)#生成區間[a, b)內的隨機數def rand(self, a, b):    return (b - a) * random.random() + a    # 生成大小 I*J 的矩陣，默認零矩陣def makeMatrix(self, I, J, fill=0.0):    m = []    for i in range(I):    m.append([fill] * J)    return m   # 函數 sigmoid，bp神經網絡前向傳播的激活函數def sigmoid(self, x):return 1.0 / (1.0 + math.exp(-x))       # 函數 sigmoid 的導數，反向傳播時使用def dsigmoid(self, x):return x * (1 - x)

3.讀取數據并進行預處理

主要是讀取構建分類模型的數據，并進行預處理。

  data = []                            raw = pd.read_csv('iris.csv')    raw_data = raw.values            raw_feature = raw_data[1:, 1:5]    for i in range(len(raw_feature)):          ele = []                    ele.append(list(raw_feature[i]))  if raw_data[i][5] == 0:   ele.append([0, 0,1])    elif raw_data[i][5] == 1:ele.append([0,1, 0])elif raw_data[i][5] == 2:ele.append([1, 1,1])else:ele.append([0, 0,0])data.append(ele)

4.利用構建的BP神經網絡預測類，創建神經網絡模型

主要是用BP神經網絡預測類創建神經網絡類模型。

  nn = NN(4, 10, 3)  

5.BP分類模型訓練及預測

主要是劃分訓練集和測試集，并進行BP分類模型訓練和預測。

   training = data[1:100]            test = data[101:]            nn.train(training, iterations=1000)  nn.test(test) 

完整源代碼

from __future__ import division    
import math      
import random    
import pandas as pd    
import numpy as np""" 三層反向傳播神經網絡 """
class NN:def __init__(self, ni, nh, no):self.ni = ni + 1                            # 輸入層節點self.nh = nh + 1                    # 隱藏層節點self.no = no                      # 輸出層種類self.ai = [1.0] * self.ni    self.ah = [1.0] * self.nh    self.ao = [1.0] * self.no    self.wi = self.makeMatrix(self.ni, self.nh)  # 輸出層到隱藏層的映射矩陣self.wo = self.makeMatrix(self.nh, self.no)  # 隱藏層到輸出層的映射矩陣for i in range(self.ni):          for j in range(self.nh):    self.wi[i][j] = self.rand(-0.2, 0.2)  for j in range(self.nh):for k in range(self.no):self.wo[j][k] = self.rand(-2, 2)  #前向傳播，激活神經網絡的所有節點def update(self, inputs):if len(inputs) != self.ni - 1:print(len(inputs),self.ni - 1)raise ValueError('與輸入層節點數不符！')    for i in range(self.ni - 1):    self.ai[i] = inputs[i]    for j in range(self.nh):                  # self.nh表示隱藏層的節點數sum = 0.0                            # 激活項a = g(z)  z = Θ^T x ;sum相當于z，每次循環歸零for i in range(self.ni):                  #通過循環z = Θ^T x ，因為Θ、x均為向量sum = sum + self.ai[i] * self.wi[i][j]  #〖 Z〗^((2))=Θ^((1)) a^((1))self.ah[j] = self.sigmoid(sum)    # a^((2))=g(z^((2)))，這里使用sigmoid()函數作為激活函數for k in range(self.no):sum = 0.0for j in range(self.nh):sum = sum + self.ah[j] * self.wo[j][k]  #〖 Z〗^((3))=Θ^((2)) a^((2))self.ao[k] = self.sigmoid(sum)    # a^((3))=g(z^((3)))return self.ao[:]#反向傳播,計算節點激活項的誤差def backPropagate(self, targets, lr):               # targets為某樣本實際種類分類，lr為梯度下降算法的學習率output_deltas = [0.0] * self.nofor k in range(self.no):error = targets[k] - np.round_(self.ao[k])output_deltas[k] = self.dsigmoid(self.ao[k]) * error# 計算隱藏層的誤差hidden_deltas = [0.0] * self.nh    for j in range(self.nh):error = 0.0for k in range(self.no):error = error + output_deltas[k] * self.wo[j][k]    hidden_deltas[j] = self.dsigmoid(self.ah[j]) * error# 更新輸出層權重for j in range(self.nh):            # 反向傳播算法，求出每個節點的誤差后，反向更新權重for k in range(self.no):change = output_deltas[k] * self.ah[j]    self.wo[j][k] = self.wo[j][k] + lr * change   # 更新輸入層權重for i in range(self.ni):                    for j in range(self.nh):change = hidden_deltas[j] * self.ai[i]self.wi[i][j] = self.wi[i][j] + lr * change# 計算誤差error = 0.0for k in range(self.no):                                    error += 0.5 * (targets[k] - np.round_(self.ao[k])) ** 2  return error                                          #用測試集輸出準確率def test(self, patterns):                            count = 0num=0for p in patterns:target = p[1]result = self.update(p[0])                    print(p[0], ':', target, '->', np.round_(result))num=0for k in range(self.no):if (target[k] == np.round_(result[k])):num +=1print(num)if num==3:count +=1print("******************",(target) == (np.round_(result)),"******************")accuracy = int(float(count / len(patterns))*100)print('accuracy: %-.9f' % accuracy,"%")      #輸出訓練過后神經網絡的權重矩陣def weights(self):print('輸入層權重:')for i in range(self.ni):print(self.wi[i])print()print('輸出層權重:')for j in range(self.nh):print(self.wo[j])#用訓練集訓練神經網絡def train(self, patterns, iterations=1000, lr=0.1):  for i in range(iterations):error = 0.0                    for p in patterns:            inputs = p[0]            targets = p[1]            self.update(inputs)          error = error + self.backPropagate(targets, lr)  if i % 100 == 0:print("percent：",int(i/iterations*100),"%",'   error: %-.9f' % error)#生成區間[a, b)內的隨機數def rand(self, a, b):    return (b - a) * random.random() + a    # 生成大小 I*J 的矩陣，默認零矩陣def makeMatrix(self, I, J, fill=0.0):    m = []    for i in range(I):    m.append([fill] * J)    return m   # 函數 sigmoid，bp神經網絡前向傳播的激活函數def sigmoid(self, x):return 1.0 / (1.0 + math.exp(-x))       # 函數 sigmoid 的導數，反向傳播時使用def dsigmoid(self, x):return x * (1 - x)if __name__ == '__main__':data = []                            raw = pd.read_csv('iris.csv')    raw_data = raw.values            raw_feature = raw_data[1:, 1:5]    for i in range(len(raw_feature)):          ele = []                    ele.append(list(raw_feature[i]))  if raw_data[i][5] == 0:   ele.append([0, 0,1])    elif raw_data[i][5] == 1:ele.append([0,1, 0])elif raw_data[i][5] == 2:ele.append([1, 1,1])else:ele.append([0, 0,0])data.append(ele)nn = NN(4, 10, 3)  training = data[1:100]            test = data[101:]            nn.train(training, iterations=1000)  nn.test(test)

?

?