BP算法鸢尾花分类

网上很多鸢尾花例子，学习其他人后仿写，我在运行其他人的时候会有溢出和错误。

下述代码准确率95%

附鸢尾花数据集链接：https://pan.baidu.com/s/1J3-2INUT19BVZWCUet3X6A

提取码：y07d

如果有什么不对的或者有什么不懂Iamzhubaoliang@yeah.net

工程地址https://git.lug.ustc.edu.cn/Iamzhubaoliang/irissort

import math

import random

import pandas as pd

random.seed(0)

def rand(a,b):

return (b-a)*random.random()+a

def dsigmoid(x):

return x*(1-x)

def sigmoid(gamma):

if gamma < 0:

return 1 - 1/(1 + math.exp(gamma))

else:

return 1/(1 + math.exp(-gamma))

def makematrix(I,J):

m = []

for i in range(I):

m.append([0.0]*J)#一维构造二维

return m

class BP:

''' 三层BP网络'''

def __init__(self,ni,nh,no):

self.ni=ni+1

self.nh=nh+1

self.no=no

self.ai=[1.0]*self.ni

self.ah=[1.0]*self.nh

self.ao=[1.0]*self.no

self.wi=makematrix(self.ni,self.nh)

self.wo=makematrix(self.nh,self.no)

for i in range(self.ni):

for j in range(self.nh):

self.wi[i][j]=rand(-2,2)

for i in range(self.nh):

for j in range(self.no):

self.wo[i][j]=rand(-2,2)

def Forward(self,input):

for i in range(self.ni-1):

self.ai=input

for i in range(self.nh):

sum=0.0

for j in range(self.ni-1):

#print(self.ai[j])

sum+=self.wi[j][i]*self.ai[j]

self.ah[i]=sigmoid(sum)

for i in range(self.no):

sum=0.0

for j in range(self.no):

sum+=self.wo[j][i]*self.ah[j]

self.ao[i]=sigmoid(sum)

return self.ao

def backward(self,tartget,lv):

rerror=0.0

outerror=[1.0]*self.no

for i in range(self.no):

error=tartget[i]-self.ao[i]

rerror=error

outerror[i]=dsigmoid(self.ao[i])*error

herror=[1.0]*self.nh

for i in range(self.nh):

error=0.0

for j in range(self.no):

error+=outerror[j]*self.wo[i][j]

herror[i]=dsigmoid(self.ah[i])*error

for i in range(self.nh):

for j in range(self.no):

change=outerror[j]*self.ah[i]

self.wo[i][j]=self.wo[i][j]+lv*change

for i in range(self.ni-1):

for j in range(self.nh):

change=herror[j]*self.ai[i]

self.wi[i][j]=self.wi[i][j]+lv*change

error = 0.0

error += 0.5 * rerror ** 2

return error

def train(self, patterns, iterations, lr=0.009):

# lr: 学习速率(learning rate)

for i in range(iterations):

error = 0.0

for p in patterns:

inputs = p[0]

targets = p[1]

#print(inputs, ",", p[1])

self.Forward(inputs)

error = error + self.backward(targets, lr)

if i%100==0:

print('error: %-.9f' % error)

def test(self,patterns):

count=0

for p in patterns:

target=p[1]

#print(target)

result=self.Forward(p[0])

index = result.index(max(result))

if index==0 and target==[1,0,0]:

count=count+1

elif index==1 and target==[0,1,0]:

count=count+1

elif index==2 and target==[0,0,1]:

count = count + 1

accuracy = float(count / len(patterns))

print('accuracy: %-.9f' % accuracy)

def iris():

data=[]

raw=pd.read_csv('iris.csv')

raw_data=raw.values

raw_feature=raw_data[:,1:6]

print(raw_feature)

for i in range(len(raw_feature)):

ele=[]

ele.append(raw_feature[i])

if raw_data[i][5]=='setosa':

ele.append([1,0,0])

elif raw_data[i][5]=='versicolor':

ele.append([0,1,0])

else:

ele.append([0,0,1])

print(ele)

data.append(ele)

random.shuffle(data)

trainning=data[0:100]

test=data[101:]

nn=BP(4,7,3)

nn.train(data,10000)

nn.test(test)

if __name__ == '__main__':

iris()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

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141