-
Notifications
You must be signed in to change notification settings - Fork 6
/
model.py
68 lines (53 loc) · 1.61 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
import torch
from torch import nn
from torch.nn import Parameter
class AutoEncoder(nn.Module):
def __init__(self, hidden, input_size):
super(AutoEncoder, self).__init__()
self.encoder = nn.Sequential(
nn.Linear(input_size, 500),
nn.ReLU(True),
nn.Linear(500, 500),
nn.ReLU(True),
nn.Linear(500, 500),
nn.ReLU(True),
nn.Linear(500, 2000),
nn.ReLU(True),
nn.Linear(2000, hidden))
self.decoder = nn.Sequential(
nn.Linear(hidden, 2000),
nn.ReLU(True),
nn.Linear(2000, 500),
nn.ReLU(True),
nn.Linear(500, 500),
nn.ReLU(True),
nn.Linear(500, 500),
nn.ReLU(True),
nn.Linear(500, input_size))
self.model = nn.Sequential(self.encoder, self.decoder)
def encode(self, x):
return self.encoder(x)
def decode(self, x):
return self.decoder(x)
def forward(self, x):
return self.model(x)
class GCML(nn.Module):
def __init__(self, n_clusters, hidden, autoencoder, alpha=1.0):
super(GCML, self).__init__()
self.n_clusters = n_clusters
self.alpha = alpha
self.autoencoder = autoencoder
self.cluster_centers = Parameter(torch.Tensor(n_clusters, hidden))
torch.nn.init.xavier_normal_(self.cluster_centers)
def target_distribution(self, q):
weight = (q ** 2.0) / torch.sum(q, 0)
return (weight.t() / torch.sum(weight, 1)).t()
def forward(self, x):
hidden = self.autoencoder.encode(x)
x_rec = self.autoencoder.decode(hidden)
p_squared = torch.sum((hidden.unsqueeze(1) - self.cluster_centers)**2, 2)
p = 1.0 / (1.0 + (p_squared / self.alpha))
power = float(self.alpha + 1) / 2
p = p ** power
p_dist = (p.t() / torch.sum(p, 1)).t()
return x_rec, p_dist