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
| from pybrain.structure import RecurrentNetwork, FullConnection, LinearLayer, SigmoidLayer
from pybrain.datasets import SupervisedDataSet
from pybrain.supervised.trainers import BackpropTrainer
def evaluate(net , X_test, Y_test):
"""
Оценка набора в процентах.
Проверяет по всему набору данных и ответов.
X_test: матрица обучающего набора X
Y_test: матрица ответов Y
return точность в процентах
"""
scores = []
res_acc = 0
rows = len(X_test)
wi_y_test = len(Y_test[0])
elem_of_out_nn = 0
elem_answer = 0
is_vecs_are_equal = False
out_nn=None
for row in range(rows):
x_test = X_test[row]
y_test = Y_test[row]
out_nn = net.activate(x_test)
print('x test', x_test)
print('out nn evaluate', out_nn)
for elem in range(wi_y_test):
elem_of_out_nn = out_nn[elem]
elem_answer = y_test[elem]
if elem_of_out_nn > 0.5:
elem_of_out_nn = 1
print("output vector elem -> ( %f ) " % 1, end=' ')
print("expected vector elem -> ( %f )" %
elem_answer, end=' ')
else:
elem_of_out_nn = 0
print("output vector elem -> ( %f ) " % 0, end=' ')
print("expected vector elem -> ( %f )" %
elem_answer, end=' ')
if elem_of_out_nn == elem_answer:
is_vecs_are_equal = True
else:
is_vecs_are_equal = False
break
if is_vecs_are_equal:
print("-Vecs are equal-")
scores.append(1)
else:
print("-Vecs are not equal-")
scores.append(0)
res_acc = sum(scores) / rows * 100
return res_acc
def activ(net, x):
out_nn=net.activate(x)
return out_nn
#Define network structure
network = RecurrentNetwork(name="XOR")
inputLayer = LinearLayer(2, name="Input")
hiddenLayer = SigmoidLayer(3, name="Hidden")
outputLayer = LinearLayer(1, name="Output")
network.addInputModule(inputLayer)
network.addModule(hiddenLayer)
network.addOutputModule(outputLayer)
c1 = FullConnection(inputLayer, hiddenLayer, name="Input_to_Hidden")
c2 = FullConnection(hiddenLayer, outputLayer, name="Hidden_to_Output")
c3 = FullConnection(hiddenLayer, hiddenLayer, name="Recurrent_Connection")
network.addConnection(c1)
network.addRecurrentConnection(c3)
network.addConnection(c2)
network.sortModules()
#Add a data set
ds = SupervisedDataSet(2,1)
X=[(0,0),
(0,1),
(1,0),
(1,1)]
Y=[(0,),
(1,),
(1,),
(0,)]
height_X_Y=len(X)
for row in range(height_X_Y):
ds.addSample(X[row], Y[row])
#Train the network
trainer = BackpropTrainer(network, ds, momentum=0.99)
max_error = 1e-7
error, count = 1, 1000
#Train
while abs(error) >= max_error and count > 0:
error = trainer.train()
count = count - 1
print("Error: ", error)
print(activ(network, [0, 0]))
print(activ(network, [0, 1]))
print(activ(network, [1, 0]))
print(activ(network, [1, 1]))
print(evaluate(network, X, Y))
print('-----------')
print(evaluate(network, X, Y)) |
