-
Notifications
You must be signed in to change notification settings - Fork 8
/
deep_and_cross_tf.py
172 lines (139 loc) · 7.04 KB
/
deep_and_cross_tf.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
172
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Created on Sat Mar 17 21:29:21 2018
@author: flyaway
"""
from sklearn.preprocessing import OneHotEncoder,StandardScaler
import numpy as np
from sklearn.metrics import accuracy_score
from sklearn.cross_validation import train_test_split
import tensorflow as tf
import pandas as pd
import keras.backend as K
def preprocessing(data):
# inverse transform one-hot to continuous column
df_onehot = data[[col for col in data.columns.tolist() if "Soil_Type" in col]]
#for i in df_onehot.columns.tolist():
# if df_onehot[i].sum() == 0:
# del df_onehot[i]
data["Soil"] = df_onehot.dot(np.array(range(df_onehot.columns.size))).astype(int)
data.drop([col for col in data.columns.tolist() if "Soil_Type" in col], axis = 1, inplace = True)
label = np.array(OneHotEncoder().fit_transform(data["Cover_Type"].values.reshape(-1, 1)).todense())
del data["Cover_Type"]
cate_columns = ["Soil"]
cont_columns = [col for col in data.columns if col != "Soil"]
# Feature normilization
scaler = StandardScaler()
data_cont = pd.DataFrame(scaler.fit_transform(data[cont_columns]), columns = cont_columns)
data_cate = data[cate_columns]
data = pd.concat([data_cate, data_cont], axis = 1)
cate_values = data[cate_columns].values
cont_values = data[cont_columns].values
return data, label, cate_values, cont_values #cate_columns, cont_columns
def embedding_input(inp,name, n_in, n_out):
#inp = Input(shape = (1, ), dtype = 'int64', name = name)
#inp = tf.placeholder(tf.int32,shape = (1,))
print type(inp)
with tf.variable_scope(name) as scope:
embeddings = tf.Variable(tf.random_uniform([n_in,n_out],-1.0,1.0))
#embeddings = tf.Variable(tf.random_uniform([voc_size, embedding_size], -1.0, 1.0))
print embeddings.shape
return inp,tf.nn.embedding_lookup(embeddings, inp,name = scope.name)
def embedding_feature_generate(cate_values,num_cate):
#data, label, cate_columns, cont_columns = preprocessing(data)
embeddings_tensors = []
col = cate_values.shape[1]
for i in range(col):
layer_name = 'inp_' + str(i)
#nunique = np.unique(cate_values[:,i]).shape[0]
nunique = num_cate[i]
embed_dim = nunique if int(6 * np.power(nunique, 1/4)) > nunique \
else int(6 * np.power(nunique, 1/4))
t_inp, t_build = embedding_input(cate_values[:,i],layer_name, nunique, embed_dim)
embeddings_tensors.append((t_inp, t_build))
del(t_inp, t_build)
inp_embed = [et[1] for et in embeddings_tensors]
return inp_embed
def fclayer(x,output_dim,reluFlag,name):
with tf.variable_scope(name) as scope:
input_dim = x.get_shape().as_list()[1]
W = tf.Variable(tf.random_normal([input_dim,output_dim], stddev=0.01))
b = tf.Variable(tf.random_normal([output_dim], stddev=0.01))
out = tf.nn.xw_plus_b(x,W,b,name = scope.name)
if reluFlag:
return tf.nn.relu(out)
else:
return out
def crosslayer(x,inp_embed,name):
with tf.variable_scope(name) as scope:
input_dim = x.get_shape().as_list()[1]
w = tf.Variable(tf.random_normal([1, input_dim], stddev=0.01))
b = tf.Variable(tf.random_normal([1,input_dim], stddev=0.01))
tmp1 = K.batch_dot(K.reshape(x, (-1, input_dim, 1)), tf.reshape(inp_embed,(-1,1,input_dim)))
tmp = K.sum(w * tmp1, 1, keepdims = True)
tmp = tf.reshape(tmp,shape=(-1,input_dim))
#one = tf.ones_like(tmp)
#bb = tf.tensordot(one ,b,1)
output = tf.add(tmp,b)
output = tf.add(output,inp_embed)
return output
def build_model(X_cate,X_cont,num_cate):
inp_embed = embedding_feature_generate(X_cate,num_cate)
inp_embed = tf.concat([inp_embed],axis = 1,name = "concat")
input_dim = inp_embed.get_shape().as_list()
inp_embed = tf.reshape(inp_embed,input_dim[1:3])
inp_embed = tf.concat([inp_embed,X_cont],axis = 1,name = "concat")
fc1 = fclayer(inp_embed,272,reluFlag=False,name = 'fc_1')
fc2 = fclayer(fc1,272,reluFlag=False,name = 'fc_2')
fc3 = fclayer(fc2,272,reluFlag=False,name = 'fc_3')
fc4 = fclayer(fc3,272,reluFlag=False,name = 'fc_4')
fc5 = fclayer(fc4,272,reluFlag=False,name = 'fc_5')
fc6 = fclayer(fc5,272,reluFlag=False,name = 'fc_5')
#return fc6
#print inp_embed.shape
cross1 = crosslayer(inp_embed,inp_embed,name = 'cross1')
#print cross1.shape
cross2 = crosslayer(cross1,inp_embed,name = 'cross2')
cross3 = crosslayer(cross2,inp_embed,name = 'cross3')
cross4 = crosslayer(cross3,inp_embed,name = 'cross4')
cross5 = crosslayer(cross4,inp_embed,name = 'cross5')
cross6 = crosslayer(cross5,inp_embed,name = 'cross6')
cross7 = crosslayer(cross6,inp_embed,name = 'cross7')
cross8 = crosslayer(cross7,inp_embed,name = 'cross8')
output = tf.concat([fc6, cross8], axis = 1,name = 'concat')
#print output.shape
out = fclayer(output,7,reluFlag=False,name = 'out')
return out
if __name__ == "__main__":
data = pd.read_csv("./covtype.csv")
data, label, cate_values, cont_values = preprocessing(data)
#获取每种类别特征的最大值，这里仅有一种类别
num_cate = []
for i in range(cate_values.shape[1]):
nunique = np.unique(cate_values[:,i]).shape[0]
num_cate.append(nunique)
X_cate = tf.placeholder(tf.int32, [None, cate_values.shape[1]])
X_cont = tf.placeholder("float",[None,cont_values.shape[1]])
Y = tf.placeholder("float", [None, 7])
#print(inp_embed.get_shape().as_list())
#input_dim = inp_embed.get_shape().as_list()
#X_train, X_test, y_train, y_test = train_test_split(inp_embed, label, test_size=0.1,random_state=1024)
output = build_model(X_cate,X_cont,num_cate)
entropy=tf.nn.softmax_cross_entropy_with_logits(logits=output,labels=Y)
cost = tf.reduce_sum(entropy)
train_op = tf.train.AdamOptimizer(0.001).minimize(cost)
predict_op = tf.arg_max(output,1)
## prepare for data
X_train_cate, X_test_cate,X_train_cont,X_test_cont,\
y_train, y_test = train_test_split(cate_values,cont_values,label, test_size=0.1,random_state=1024)
with tf.Session() as sess:
sess.run(tf.global_variables_initializer())
for i in range(100):
lost = []
for start, end in zip(range(0, len(X_train_cate), 256), range(256, len(X_train_cate)+1, 256)):
sess.run(train_op, feed_dict={X_cate: X_train_cate[start:end], X_cont: X_train_cont[start:end],Y:y_train[start:end]})
lost.append(sess.run(cost,feed_dict={X_cate: X_train_cate[start:end], X_cont: X_train_cont[start:end],Y:y_train[start:end]}))
print(i,np.mean(lost))
print(i,np.mean(np.argmax(y_train, axis=1) == sess.run(predict_op, feed_dict={X_cate: X_train_cate[start:end], X_cont: X_train_cont[start:end]})))
print(i, np.mean(np.argmax(y_test, axis=1) == sess.run(predict_op, feed_dict={X_cate: X_test_cate[start:end], X_cont: X_test_cont[start:end]})))