deeppacket/model.py at master · KimythAnly/deeppacket · GitHub

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
import keras
from keras.models import Sequential
from keras.layers import Dense, Dropout, Conv1D, MaxPooling1D, Flatten, Activation

def encode_by_ae(ae, data, batch_size = 32, verbose = True):
    model = Sequential()
    model.add(ae.get_encoder())
    return model.predict(data, batch_size = batch_size, verbose = verbose)


class AutoEncoder(Sequential):
    def __init__(self, input_size, encode_size, dropout = 0.05, encoder_id = 0):
        super(AutoEncoder, self).__init__()
        self.params = { 'input_size': input_size,
                        'encode_size': encode_size,
                        'dropout': dropout,
                        'encoder_id': encoder_id}
        self.encoder_id = encoder_id
        self.add(Dense(encode_size, input_shape = (input_size,), activation = 'relu', name = 'encoder{}'.format(encoder_id)))
        self.add(Dropout(dropout))
        self.add(Dense(input_size, activation = 'sigmoid', name = 'decoder{}'.format(encoder_id)))
        self.add(Dropout(dropout))

    def get_encoder(self):
        return self.get_layer('encoder{}'.format(self.encoder_id))

    def get_decoder(self):
        return self.get_layer('decoder{}'.format(self.encoder_id))

    def get_config(self):
        config = super().get_config()
        for key in self.params.keys():
            config[key] = self.params[key]
        return config

class StackedAutoEncoder(Sequential):
    def __init__(self, auto_encoders, dropout = 0.05):
        super(StackedAutoEncoder, self).__init__()
        self.params = {'auto_encoders': auto_encoders, 'dropout': dropout}
        self.encoder_layers = [ae.get_encoder() for ae in auto_encoders]
        self.decoder_layers = reversed([ae.get_decoder() for ae in auto_encoders])
        for i, e in enumerate(self.encoder_layers):
            if i > 0:
                self.add(Dropout(dropout))
            self.add(e)
        for d in self.decoder_layers:
            self.add(Dropout(dropout))
            self.add(d)
    def get_encoder(self):
        return self.encoder_layers
    def get_config(self):
        config = super().get_config()
        for key in self.params.keys():
            config[key] = self.params[key]
        return config

class StackedAutoEncoderClassifier(Sequential):
    def __init__(self, stacked_auto_encoder, dropout = 0.05):
        super(StackedAutoEncoderClassifier, self).__init__()
        self.params = {'stacked_auto_encoder':stacked_auto_encoder, 'dropout': dropout}
        self.stacked_auto_encoder = stacked_auto_encoder
        for encoder in stacked_auto_encoder.get_encoder():
            self.add(encoder)
            self.add(Dropout(dropout))
        self.add(Dense(17, activation = 'softmax'))

    def get_config(self):
        config = super().get_config()
        for key in self.params.keys():
            config[key] = self.params[key]
        return config


class CNN(Sequential):
    def __init__(self, input_size = 1500, dropout = 0.05):
        super(CNN, self).__init__()
        self.params = {'input_size': input_size, 'dropout': dropout}
        self.add(Conv1D(200, 5, input_shape = (input_size,1), activation = 'relu'))
        self.add(Dropout(dropout))
        self.add(Conv1D(100, 4, activation = 'relu'))
        self.add(Dropout(dropout))
        self.add(MaxPooling1D(2))
        self.add(Flatten())
        denses = [600, 500, 400, 300, 200, 100, 50]
        for dense in denses:
            self.add(Dense(dense, activation = 'relu'))
            self.add(Dropout(dropout))
        self.add(Dense(17, activation = 'softmax'))

    def get_config(self):
        config = super().get_config()
        for key in self.params.keys():
            config[key] = self.params[key]
        return config


class CNN2(Sequential):
    def __init__(self, input_size = 1500, dropout = 0.05):
        super(CNN2, self).__init__()
        self.params = {'input_size': input_size, 'dropout': dropout}
        self.add(Conv1D(200, 5, input_shape = (input_size,1), activation = 'relu'))
        self.add(Dropout(dropout))
        self.add(Conv1D(100, 4, activation = 'relu'))
        self.add(Dropout(dropout))
        self.add(MaxPooling1D(2))
        self.add(Flatten())
        denses = [600, 500, 400, 300, 200, 100, 50]
        for dense in denses:
            self.add(Dense(dense, activation = 'relu'))
            self.add(Dropout(dropout))
        self.add(Dense(12, activation = 'softmax'))

    def get_config(self):
        config = super().get_config()
        for key in self.params.keys():
            config[key] = self.params[key]
        return config

class StackedAutoEncoderClassifier2(Sequential):
    def __init__(self, stacked_auto_encoder, dropout = 0.05):
        super(StackedAutoEncoderClassifier2, self).__init__()
        self.params = {'stacked_auto_encoder':stacked_auto_encoder, 'dropout': dropout}
        self.stacked_auto_encoder = stacked_auto_encoder
        for encoder in stacked_auto_encoder.get_encoder():
            self.add(encoder)
            self.add(Dropout(dropout))
        self.add(Dense(12, activation = 'softmax'))

    def get_config(self):
        config = super().get_config()
        for key in self.params.keys():
            config[key] = self.params[key]
        return config