-
Notifications
You must be signed in to change notification settings - Fork 1
/
model.py
171 lines (142 loc) · 5.75 KB
/
model.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
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
import torch
import torch.nn as nn
from torch.nn import functional as F
from utils import sparse_dropout
class LSTMnet(nn.Module):
def __init__(self,in_dim,hidden_dim,n_layer,n_class):
super(LSTMnet,self).__init__()
self.n_layer = n_layer
self.hidden_dim = hidden_dim
self.lstm = nn.LSTM(in_dim, hidden_dim, n_layer, batch_first=True)
self.linear = nn.Linear(hidden_dim, n_class)
self.sigmoid = nn.Sigmoid()
def forward(self,x):
out,_=self.lstm(x)
out = out[:,-1,:]
out=self.linear(out)
out=torch.softmax(out,dim=1)
return out
class ANNnet(nn.Module):
def __init__(self,in_dim,hidden_dim,n_layer,n_class):
super(ANNnet,self).__init__()
self.in_dim=in_dim
self.hidden_dim=hidden_dim
self.n_layer=n_layer
self.n_class=n_class
self.fc1=nn.Linear(in_features=in_dim,out_features=hidden_dim)
self.relu=nn.ReLU()
layers=[]
for i in range(1,n_layer):
# layers.append(nn.Dropout(p=0.5))
layers.append(nn.Linear(in_features=hidden_dim,out_features=hidden_dim))
layers.append(nn.ReLU())
self.layers=nn.Sequential(*layers)
self.classifier=nn.Linear(in_features=hidden_dim,out_features=n_class)
def forward(self,x):
x = self.relu(self.fc1(x))#[B,T,F] -> [B,T,H_dim]
x = self.layers(x) #[B,T,H_dim] -> [B,T,H_dim]
x = torch.mean(x,dim=1) #[B,H_dim]
x = self.classifier(x) #[B,H_dim] -> [B,C]
out=torch.softmax(x,dim=1)
return out
class CNNnet(nn.Module):
def __init__(self,in_dim,hidden_dim,n_layer,n_class):
super(CNNnet,self).__init__()
self.in_dim=in_dim
self.hidden_dim=hidden_dim
self.n_layer=n_layer
self.n_class=n_class
self.conv1=nn.Conv1d(in_channels=self.in_dim,out_channels=self.hidden_dim,kernel_size=3,stride=1,padding=1)
self.relu=nn.ReLU()
layers=[]
for i in range(1,n_layer):
layers.append(nn.Conv1d(in_channels=self.hidden_dim,out_channels=self.hidden_dim,kernel_size=3,stride=1,padding=1))
layers.append(nn.ReLU())
self.layers=nn.Sequential(*layers)
self.classifier=nn.Conv1d(in_channels=self.hidden_dim, out_channels=self.n_class, kernel_size=1, stride=1, padding=0, bias=False)
self.sigmoid=nn.Sigmoid()
def forward(self,x):
x = x.permute(0,2,1)#[B,T,F] -> [B,F,T]
x = self.relu(self.conv1(x))#[B,F,T] -> [B,H_dim,T]
x = self.layers(x)
x = self.classifier(x)#[B,H_dim,T] -> [B,C,T]
x = x.permute(0,2,1)#[B,C,T] -> [B,T,C]
x = self.sigmoid(x)
out = torch.mean(x, dim=1)
return out
class GCNnet(nn.Module):
def __init__(self, input_dim, hidden_dim, output_dim,num_channel):
super(GCNnet, self).__init__()
self.input_dim = input_dim #L_win
self.output_dim = output_dim #n_class
self.dropout=0
print('input dim:', input_dim)
print('output dim:', output_dim)
self.gclayer = GraphConvolution(self.input_dim, 1024,
activation=F.relu,
dropout=self.dropout,
is_sparse_inputs=False)
self.mlp=nn.Sequential(
nn.Linear(in_features=num_channel,out_features=hidden_dim),
nn.ReLU(),
nn.Linear(in_features=hidden_dim,out_features=hidden_dim),
nn.ReLU()
)
self.classifier=nn.Linear(in_features=hidden_dim,out_features=output_dim)
def forward(self, x, support):
x=x.permute(0,2,1) #[B,T,C] -> [B,C,T]
x = self.gclayer((x.to(torch.float32), support))#[B,C,1024]
x=x[0].permute(0,2,1)#[B,C,1024] -> [B,1024,C]
x=self.mlp(x) #[B,1024,C] -> [B,1024,H_dim]
x = torch.mean(x,dim=1) #[B,1024,H_dim] -> [B,H_dim]
x=self.classifier(x) #[B,H_dim] -> [B,n_class]
out=torch.softmax(x,dim=1)
return out
def l2_loss(self):
layer = self.layers.children()
layer = next(iter(layer))
loss = None
for p in layer.parameters():
if loss is None:
loss = p.pow(2).sum()
else:
loss += p.pow(2).sum()
return loss
class GraphConvolution(nn.Module):
def __init__(self, input_dim, output_dim,
dropout=0.,
is_sparse_inputs=False,
bias=False,
activation = F.relu,
featureless=False):
super(GraphConvolution, self).__init__()
self.dropout = dropout
self.bias = bias
self.activation = activation
self.is_sparse_inputs = is_sparse_inputs
self.featureless = featureless
self.weight = nn.Parameter(torch.randn(input_dim, output_dim))
self.bias = None
if bias:
self.bias = nn.Parameter(torch.zeros(output_dim))
def forward(self, inputs):
x, support = inputs
if self.training and self.is_sparse_inputs:
num_features_nonzero=x._nnz()
x = sparse_dropout(x, self.dropout, num_features_nonzero)
elif self.training:
x = F.dropout(x, self.dropout)
# convolve
if not self.featureless: # if it has features x
if self.is_sparse_inputs:
xw = torch.sparse.mm(x, self.weight)
else:
weight=self.weight.repeat(x.shape[0],1,1)
xw = torch.bmm(x, weight)
else:
xw = self.weight
b_support=support.repeat(x.shape[0],1,1)
out=torch.bmm(b_support,xw)
if self.bias is not None:
out += self.bias
return self.activation(out), support