-
Notifications
You must be signed in to change notification settings - Fork 95
/
VaDE_test_reuters_all.py
163 lines (144 loc) · 5.55 KB
/
VaDE_test_reuters_all.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
# -*- coding: utf-8 -*-
"""
VaDE (Variational Deep Embedding:A Generative Approach to Clustering)
Reuters(685K) clustering accuracy: 79.38%
@code author: Zhuxi Jiang
"""
import numpy as np
from keras.layers import Input, Dense, Lambda
from keras.models import Model
from keras import backend as K
import scipy.io as scio
import theano
import theano.tensor as T
import math
from sklearn import preprocessing
from keras.layers.core import Dropout,Activation
import copy
import warnings
warnings.filterwarnings("ignore")
def floatX(X):
return np.asarray(X, dtype=theano.config.floatX)
def sampling(args):
z_mean, z_log_var = args
epsilon = K.random_normal(shape=(batch_size, latent_dim), mean=0.)
return z_mean + K.exp(z_log_var / 2) * epsilon
#=====================================
def cluster_acc(Y_pred, Y):
from sklearn.utils.linear_assignment_ import linear_assignment
assert Y_pred.size == Y.size
D = max(Y_pred.max(), Y.max())+1
w = np.zeros((D,D), dtype=np.int64)
for i in range(Y_pred.size):
w[Y_pred[i], Y[i]] += 1
ind = linear_assignment(w.max() - w)
return sum([w[i,j] for i,j in ind])*1.0/Y_pred.size
#==================================================
def load_data():
print ('loading data....')
from sklearn.feature_extraction.text import CountVectorizer
did_to_cat = {}
cat_list = ['CCAT', 'GCAT', 'MCAT', 'ECAT']
with open('dataset/reuters/rcv1-v2.topics.qrels') as fin:
for line in fin.readlines():
line = line.strip().split(' ')
cat = line[0]
did = int(line[1])
if cat in cat_list:
did_to_cat[did] = did_to_cat.get(did, []) + [cat]
copy_dc=copy.copy(did_to_cat)
for did in copy_dc.keys():
if len(did_to_cat[did]) > 1:
del did_to_cat[did]
dat_list = ['lyrl2004_tokens_test_pt0.dat',
'lyrl2004_tokens_test_pt1.dat',
'lyrl2004_tokens_test_pt2.dat',
'lyrl2004_tokens_test_pt3.dat',
'lyrl2004_tokens_train.dat']
data = []
target = []
cat_to_cid = {'CCAT':0, 'GCAT':1, 'MCAT':2, 'ECAT':3}
del did
for dat in dat_list:
print (dat+'....')
with open('dataset/reuters/'+dat) as fin:
for line in fin.readlines():
if line.startswith('.I'):
if 'did' in locals():
assert doc != ''
if did in did_to_cat.keys():
data.append(doc)
target.append(cat_to_cid[did_to_cat[did][0]])
did = int(line.strip().split(' ')[1])
doc = ''
elif line.startswith('.W'):
assert doc == ''
else:
doc += line
assert len(data) == len(did_to_cat)
X = CountVectorizer(dtype=np.float64, max_features=2000).fit_transform(data)
Y = np.asarray(target)
from sklearn.feature_extraction.text import TfidfTransformer
X = TfidfTransformer(norm='l2', sublinear_tf=True).fit_transform(X)
X = np.asarray(X.todense())*np.sqrt(X.shape[1])
X = preprocessing.normalize(X, norm='l2')*200
X = X.astype('float32')
X = X[:685000] # for 100 minibatch training
Y = Y[:685000]
return X,Y
def gmm_para_init():
gmm_weights=scio.loadmat('trained_model_weights/reuters_all_weights_gmm.mat')
u_init=gmm_weights['u']
lambda_init=gmm_weights['lambda']
theta_init=np.squeeze(gmm_weights['theta'])
theta_p=theano.shared(np.asarray(theta_init,dtype=theano.config.floatX),name="pi")
u_p=theano.shared(np.asarray(u_init,dtype=theano.config.floatX),name="u")
lambda_p=theano.shared(np.asarray(lambda_init,dtype=theano.config.floatX),name="lambda")
return theta_p,u_p,lambda_p
#================================
def get_gamma(tempz):
temp_Z=T.transpose(K.repeat(tempz,n_centroid),[0,2,1])
temp_u_tensor3=T.repeat(u_p.dimshuffle('x',0,1),batch_size,axis=0)
temp_lambda_tensor3=T.repeat(lambda_p.dimshuffle('x',0,1),batch_size,axis=0)
temp_theta_tensor3=theta_p.dimshuffle('x','x',0)*T.ones((batch_size,latent_dim,n_centroid))
temp_p_c_z=K.exp(K.sum((K.log(temp_theta_tensor3)-0.5*K.log(2*math.pi*temp_lambda_tensor3)-\
K.square(temp_Z-temp_u_tensor3)/(2*temp_lambda_tensor3)),axis=1))
return temp_p_c_z/K.sum(temp_p_c_z,axis=-1,keepdims=True)
#=====================================================
ispretrain = True
batch_size = 100
latent_dim = 10
intermediate_dim = [500,500,2000]
theano.config.floatX='float32'
X,Y = load_data()
original_dim = 2000
n_centroid = 4
theta_p, u_p, lambda_p = gmm_para_init()
#===================
x = Input(batch_shape=(batch_size, original_dim))
h = Dense(intermediate_dim[0])(x)
h=Activation('relu')(h)
h=Dropout(0.2)(h)
h = Dense(intermediate_dim[1])(h)
h=Activation('relu')(h)
h=Dropout(0.2)(h)
h = Dense(intermediate_dim[2])(h)
h=Activation('relu')(h)
h=Dropout(0.2)(h)
z_mean = Dense(latent_dim)(h)
z_log_var = Dense(latent_dim)(h)
z = Lambda(sampling, output_shape=(latent_dim,))([z_mean, z_log_var])
z = Lambda(sampling, output_shape=(latent_dim,))([z_mean, z_log_var])
h_decoded = Dense(intermediate_dim[-1])(z)
h_decoded = Dense(intermediate_dim[-2])(h_decoded)
h_decoded = Dense(intermediate_dim[-3])(h_decoded)
x_decoded_mean = Dense(original_dim)(h_decoded)
#========================
p_c_z = Lambda(get_gamma, output_shape=(n_centroid,))(z_mean)
sample_output = Model(x, z_mean)
p_c_z_output = Model(x, p_c_z)
#===========================================
vade = Model(x, x_decoded_mean)
vade.load_weights('trained_model_weights/reuters_all_weights_nn.h5')
accuracy = cluster_acc(np.argmax(p_c_z_output.predict(X,batch_size=batch_size),axis=1),Y)
print ('Reuters_all dataset VaDE - clustering accuracy: %.2f%%'%(accuracy*100))