-
Notifications
You must be signed in to change notification settings - Fork 0
/
NeuralTests.py
143 lines (96 loc) · 3.34 KB
/
NeuralTests.py
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
# -*- coding: utf-8 -*-
"""
Created on Tue Jan 23 12:14:32 2018
@author: wgrant14
"""
import numpy as np
import NeuralNet as nn
aPimaRawData = np.loadtxt('pima.data', delimiter = ',')
aAllInputs = aPimaRawData[:, :8]
aAllTargets = aPimaRawData[:, 8:]
aMins = aAllInputs.min(axis = 0)
aMaxs = aAllInputs.max(axis = 0)
aSpreads = aMaxs - aMins
aAllInputsNorm = (aAllInputs - aMins) / aSpreads
tsPima = [(aAllInputsNorm[k], aAllTargets[k]) for k in range(aAllInputs.shape[0])]
# Split into actual training set, validation set, and test set.
tsPimaTrain = tsPima[0::2]
tsPimaValid = tsPima[1::4]
tsPimaTest = tsPima[3::4]
np.random.seed(10)
#nnPima = nn.NeuralNet(8, [28, 6, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#nnPima = nn.NeuralNet(8, [12, 6, 1], nn.Sigmoid, 3.0)
#iBlockSize = 200
#Finished
# Number of repetitions: 1800
# Final fraction correct: 0.4739583333333333
#nnPima = nn.NeuralNet(8, [12, 6, 1], nn.Sigmoid, 1.0)
#iBlockSize = 200
#Finished
# Number of repetitions: 400
# Final fraction correct: 0.25
#nnPima = nn.NeuralNet(8, [8, 6, 1], nn.Sigmoid, 3.0)
#iBlockSize = 200
#Finished
# Number of repetitions: 600
# Final fraction correct: 0.3541666666666667
#nnPima = nn.NeuralNet(8, [8, 6, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 1500
# Final fraction correct: 0.3958333333333333
#nnPima = nn.NeuralNet(8, [16, 6, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 2500
# Final fraction correct: 0.5833333333333334
#nnPima = nn.NeuralNet(8, [16, 8, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 3500
# Final fraction correct: 0.5625
#nnPima = nn.NeuralNet(8, [16, 4, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 3500
# Final fraction correct: 0.53125
#nnPima = nn.NeuralNet(8, [16, 8, 3], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 2000
# Final fraction correct: 0.4947916666666667
#nnPima = nn.NeuralNet(8, [16, 7, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 2000
# Final fraction correct: 0.484375
#nnPima = nn.NeuralNet(8, [16, 5, 1], nn.Sigmoid, 3.0)
#iBlockSize = 500
#Finished
# Number of repetitions: 3500
# Final fraction correct: 0.5833333333333334
nnPima = nn.NeuralNet(8, [32, 6, 1], nn.Sigmoid, 3.0)
iBlockSize = 500
#Finished
# Number of repetitions: 3500
# Final fraction correct: 0.6041666666666666
# Repeat training in blocks of 100 repetitions. Continue to train as long as
# the fraction of outputs within 0.1 of their targets increases.
iBlockCount = 0
iRepetitions = 0
dCurrFracCorrect = 0.0
bContinue = True
while bContinue:
iBlockCount += 1
print('Block number', iBlockCount, flush=True)
nnPima.vLearnManyPasses(tsPimaTrain, 0.25, iBlockSize)
iRepetitions += iBlockSize
dOldFracCorrect = dCurrFracCorrect
dCurrFracCorrect = nnPima.dShowTSPerform(tsPimaValid, 0.1)
print('Fraction correct', dCurrFracCorrect, flush=True)
bContinue = (dCurrFracCorrect > dOldFracCorrect)
# Finished. Print final results.
print('Finished')
print('\tNumber of repetitions:', iRepetitions)
print('\tFinal fraction correct:', nnPima.dShowTSPerform(tsPimaTest, 0.1))