/
Add-Cluster_centroid_and_Neural_Network.py
250 lines (220 loc) · 15.3 KB
/
Add-Cluster_centroid_and_Neural_Network.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
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
#import libraries
import numpy as np
np.random.seed(13)
import pandas as pd
from time import clock
from sklearn.cluster import KMeans as kmeans
#1. Read in XX(Dimensionality Reduction) Transformed Credit Card data
Creditcard_col = ['1','2','3','4','5','6','7','8','9','10','11','12','13','14','15','16','17','18','19','20','21','22','23']
#1.1import PCA data
PCA_CreditCard_train = pd.read_csv('Credit_card2_PCA_24_train_.csv')
PCA_CreditCard_trainX = PCA_CreditCard_train[Creditcard_col]
PCA_CreditCard_trainY = PCA_CreditCard_train['Class']
PCA_CreditCard_Test = pd.read_csv('Credit_card2_PCA_24_test_.csv')
PCA_CreditCard_TestX = PCA_CreditCard_Test[Creditcard_col]
PCA_CreditCard_TestY = PCA_CreditCard_Test['Class']
#1.2.Import ICA Data
ICA_CreditCard_train = pd.read_csv('Credit_card2_ICA_train.csv')
ICA_CreditCard_trainX = ICA_CreditCard_train[Creditcard_col]
ICA_CreditCard_trainY = ICA_CreditCard_train['Class']
ICA_CreditCard_Test = pd.read_csv('Credit_card2_ICA_test.csv')
ICA_CreditCard_TestX = ICA_CreditCard_Test[Creditcard_col]
ICA_CreditCard_TestY = ICA_CreditCard_Test['Class']
#1.3.Import Random Projection Data
RP_CreditCard_train = pd.read_csv('Credit_card2_RP_train.csv')
RP_CreditCard_trainX = RP_CreditCard_train[Creditcard_col]
RP_CreditCard_trainY = RP_CreditCard_train['Class']
RP_CreditCard_Test = pd.read_csv('Credit_card2_RP_test.csv')
RP_CreditCard_TestX = RP_CreditCard_Test[Creditcard_col]
RP_CreditCard_TestY = RP_CreditCard_Test['Class']
#1.4.Import Random Forest Data
RF_CreditCard_train = pd.read_csv('CreditCard_RF_train.csv')
RF_CreditCard_trainX = RF_CreditCard_train[Creditcard_col]
RF_CreditCard_trainY = RF_CreditCard_train['Class']
RF_CreditCard_Test = pd.read_csv('CreditCard_RF_test.csv')
RF_CreditCard_TestX = RF_CreditCard_Test[Creditcard_col]
RF_CreditCard_TestY = RF_CreditCard_Test['Class']
def clustering_append_centroid( X_data,algorithm,clusters = [ 2, 5, 10, 25]):
X_data2_ = X_data[['1','2','3','4','5','6','7','8','9','10','11','12','13','14','15','16','17','18','19','20','21','22','23']]
km = kmeans(random_state=15)
for k in clusters:
km.set_params(n_clusters=k)
km = km.fit(X_data2_)
X_data2_['Cluster'+str(k)] = km.labels_
#X_train2_.to_csv(str(algorithm)+'Transformed-Add-Cluster-2-5-10-25-Credit_card_train.csv')
return X_data2_
PCA_Credit_card_train_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=PCA_CreditCard_trainX,algorithm='PCA')
ICA_Credit_card_train_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=ICA_CreditCard_trainX,algorithm='ICA')
RP_Credit_card_train_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=RP_CreditCard_trainX,algorithm='RP')
RF_Credit_card_train_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=RF_CreditCard_trainX,algorithm='RF')
PCA_Credit_card_Test_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=PCA_CreditCard_TestX,algorithm='PCA')
ICA_Credit_card_Test_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=ICA_CreditCard_TestX,algorithm='ICA')
RP_Credit_card_Test_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=RP_CreditCard_TestX,algorithm='RP')
RF_Credit_card_Test_ADD_Cluter2_5_10_25 = clustering_append_centroid(X_data=RF_CreditCard_TestX,algorithm='RF')
# print('PCA',PCA_Credit_card_Test_ADD_Cluter2_5_10_25.head(2))
# print('ICA',ICA_Credit_card_Test_ADD_Cluter2_5_10_25.head(2))
# print('RP',RP_Credit_card_Test_ADD_Cluter2_5_10_25.head(2))
# print('RF',RF_Credit_card_Test_ADD_Cluter2_5_10_25.head(2))
#2.Neural Network Training and Training data accuracy
from keras.models import Sequential
from keras.layers import Dense
from keras.optimizers import Adam
from keras import optimizers
from keras import backend as k
from keras.utils import to_categorical
import os
import matplotlib.pyplot as plt
from sklearn.metrics import accuracy_score
from keras.layers.advanced_activations import LeakyReLU, PReLU
#Experiment - Learning CUrve
def Experiment_NN(xtrain, ytrain,xtest,ytest,
cluster_feature,Predictor_type = 'Binary',
layer1_dense = 50, hidden_layer_dense = 40,output_layer_dense = 1,learnRate = 0.001,
num_of_component= ['1','2','3','4','5','6','7','8','9','10','11','12',
'13','14','15','16','17','18','19','20','21','22','23']):
# num_of_component = [1,23]):
RMSD_List_E = []
if len(cluster_feature) > 0:
num_of_component.append(cluster_feature)
X_nk_train = xtrain[num_of_component]
X_nk_train = np.asarray(X_nk_train)
ytrain = np.asarray(ytrain)
X_nk_test = xtest[num_of_component]
X_nk_test= np.asarray(X_nk_test)
Y_nk_test = np.asarray(ytest)
input_dimension = len(num_of_component)
X_nk_train = X_nk_train[0:7000]
Y_nk_train = ytrain[0:7000]
#class_weights = {0: 1., 1: 1}
model_name = Sequential()
model_name.add(Dense(layer1_dense, input_dim=input_dimension, activation='relu'))
# model_name.add(LeakyReLU(alpha=0.003))
model_name.add(Dense(hidden_layer_dense, activation='relu'))
if Predictor_type == 'Binary':
model_name.add(Dense(1, kernel_initializer='normal', activation='sigmoid'))
adam = optimizers.Adam(lr=learnRate)
model_name.compile(loss='binary_crossentropy', optimizer=adam, metrics=['accuracy']) # 'rmsprop'
model_name.fit(X_nk_train, Y_nk_train, verbose=0, # class_weight=class_weights,
epochs=300, batch_size=20)
else:
model_name.add(Dense(output_layer_dense, kernel_initializer='normal', activation='sigmoid'))
adam = optimizers.Adam(lr=learnRate)
model_name.compile(loss='categorical_crossentropy', optimizer=adam, metrics=['accuracy'])
Y_nk_train_dummy = to_categorical(Y_nk_train)
model_name.fit(X_nk_train, Y_nk_train_dummy, verbose=0,# class_weight=class_weights,
epochs=300, batch_size=20)
y_predict_train = model_name.predict_classes(X_nk_train)
y_predict_test = model_name.predict_classes(X_nk_test)
RMSD_train = accuracy_score(y_predict_train, Y_nk_train)
RMSD_test = accuracy_score(y_predict_test, Y_nk_test)
RMSD_List_E.append([cluster_feature, round(RMSD_train, 2),round(RMSD_test,2)])
return RMSD_List_E
#2.1 PCA -Accuracy
PCA_Accuracy = []
PCA_NN_Add_cluster2 = Experiment_NN(xtrain = PCA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = PCA_CreditCard_trainY,
xtest = PCA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest = PCA_CreditCard_TestY,
cluster_feature = 'Cluster2')
PCA_NN_Add_cluster5 = Experiment_NN(xtrain = PCA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = PCA_CreditCard_trainY,
xtest=PCA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=PCA_CreditCard_TestY,
cluster_feature = 'Cluster5')
PCA_NN_Add_cluster10 = Experiment_NN(xtrain = PCA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = PCA_CreditCard_trainY,
xtest=PCA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=PCA_CreditCard_TestY,
cluster_feature = 'Cluster10')
PCA_NN_Add_cluster25 = Experiment_NN(xtrain = PCA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = PCA_CreditCard_trainY,
xtest=PCA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=PCA_CreditCard_TestY,
cluster_feature = 'Cluster25')
PCA_Accuracy.append(PCA_NN_Add_cluster2)
PCA_Accuracy.append(PCA_NN_Add_cluster5)
PCA_Accuracy.append(PCA_NN_Add_cluster10)
PCA_Accuracy.append(PCA_NN_Add_cluster25)
print('PCA-Training+Test',PCA_Accuracy)
#2.2 ICA -Accuracy
ICA_Accuracy = []
ICA_NN_Add_cluster2 = Experiment_NN(xtrain = ICA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = ICA_CreditCard_trainY,
xtest=ICA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=ICA_CreditCard_TestY,
cluster_feature = 'Cluster2')
ICA_NN_Add_cluster5 = Experiment_NN(xtrain = ICA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = ICA_CreditCard_trainY,
xtest=ICA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=ICA_CreditCard_TestY,
cluster_feature = 'Cluster5')
ICA_NN_Add_cluster10 = Experiment_NN(xtrain = ICA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = ICA_CreditCard_trainY,
xtest=ICA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=ICA_CreditCard_TestY,
cluster_feature = 'Cluster10')
ICA_NN_Add_cluster25 = Experiment_NN(xtrain = ICA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = ICA_CreditCard_trainY,
xtest=ICA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=ICA_CreditCard_TestY,
cluster_feature = 'Cluster25')
ICA_Accuracy.append(ICA_NN_Add_cluster2)
ICA_Accuracy.append(ICA_NN_Add_cluster5)
ICA_Accuracy.append(ICA_NN_Add_cluster10)
ICA_Accuracy.append(ICA_NN_Add_cluster25)
print('ICA-Training+Test',ICA_Accuracy)
#2.3 RP -Accuracy
RP_Accuracy = []
RP_NN_Add_cluster2 = Experiment_NN(xtrain = RP_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RP_CreditCard_trainY,
xtest=RP_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RP_CreditCard_TestY,
cluster_feature = 'Cluster2')
RP_NN_Add_cluster5 = Experiment_NN(xtrain = RP_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RP_CreditCard_trainY,
xtest=RP_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RP_CreditCard_TestY,
cluster_feature = 'Cluster5')
RP_NN_Add_cluster10 = Experiment_NN(xtrain = RP_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RP_CreditCard_trainY,
xtest=RP_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RP_CreditCard_TestY,
cluster_feature = 'Cluster10')
RP_NN_Add_cluster25 = Experiment_NN(xtrain = RP_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RP_CreditCard_trainY,
xtest=RP_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RP_CreditCard_TestY,
cluster_feature = 'Cluster25')
RP_Accuracy.append(RP_NN_Add_cluster2)
RP_Accuracy.append(RP_NN_Add_cluster5)
RP_Accuracy.append(RP_NN_Add_cluster10)
RP_Accuracy.append(RP_NN_Add_cluster25)
print('RP-Training+Test',RP_Accuracy)
#2.4 RF -Accuracy
RF_Accuracy = []
RF_NN_Add_cluster2 = Experiment_NN(xtrain = RF_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RF_CreditCard_trainY,
xtest=RF_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RF_CreditCard_TestY,
cluster_feature = 'Cluster2')
RF_NN_Add_cluster5 = Experiment_NN(xtrain = RF_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RF_CreditCard_trainY,
xtest=RF_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RF_CreditCard_TestY,
cluster_feature = 'Cluster5')
RF_NN_Add_cluster10 = Experiment_NN(xtrain = RF_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RF_CreditCard_trainY,
xtest=RF_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RF_CreditCard_TestY,
cluster_feature = 'Cluster10')
RF_NN_Add_cluster25 = Experiment_NN(xtrain = RF_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RF_CreditCard_trainY,
xtest=RF_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RF_CreditCard_TestY,
cluster_feature = 'Cluster25')
RF_Accuracy.append(RF_NN_Add_cluster2)
RF_Accuracy.append(RF_NN_Add_cluster5)
RF_Accuracy.append(RF_NN_Add_cluster10)
RF_Accuracy.append(RF_NN_Add_cluster25)
print('RF-Training+Test',RF_Accuracy)
#3. Use the Clustering Feature only for NN training and get the training accuracy
PCA_NN_Add_cluster2_5_10_25 = Experiment_NN(xtrain = PCA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = PCA_CreditCard_trainY,
xtest=PCA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=PCA_CreditCard_TestY,
cluster_feature = 'Cluster2',num_of_component = ['Cluster5','Cluster10','Cluster25'])
ICA_NN_Add_cluster2_5_10_25 = Experiment_NN(xtrain = ICA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = ICA_CreditCard_trainY,
xtest=ICA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=ICA_CreditCard_TestY,
cluster_feature = 'Cluster2',num_of_component = ['Cluster5','Cluster10','Cluster25'])
RP_NN_Add_cluster2_5_10_25 = Experiment_NN(xtrain = RP_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RP_CreditCard_trainY,
xtest=RP_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RP_CreditCard_TestY,
cluster_feature = 'Cluster2',num_of_component = ['Cluster5','Cluster10','Cluster25'])
RF_NN_Add_cluster2_5_10_25 = Experiment_NN(xtrain = RF_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RF_CreditCard_trainY,
xtest=RF_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RF_CreditCard_TestY,
cluster_feature = 'Cluster2',num_of_component = ['Cluster5','Cluster10','Cluster25'])
print('PCA_NN_Add_cluster2_5_10_25',PCA_NN_Add_cluster2_5_10_25)
print('ICA_NN_Add_cluster2_5_10_25',ICA_NN_Add_cluster2_5_10_25)
print('RP_NN_Add_cluster2_5_10_25',RP_NN_Add_cluster2_5_10_25)
print('RF_NN_Add_cluster2_5_10_25',RF_NN_Add_cluster2_5_10_25)
PCA_NN_Add_cluster25 = Experiment_NN(xtrain = PCA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = PCA_CreditCard_trainY,
xtest=PCA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=PCA_CreditCard_TestY,
cluster_feature= [], num_of_component = ['Cluster25'])
ICA_NN_Add_cluster10 = Experiment_NN(xtrain = ICA_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = ICA_CreditCard_trainY,
xtest=ICA_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=ICA_CreditCard_TestY,
cluster_feature=[], num_of_component = ['Cluster10'])
RP_NN_Add_cluster2 = Experiment_NN(xtrain = RP_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RP_CreditCard_trainY,
xtest=RP_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RP_CreditCard_TestY,
cluster_feature=[], num_of_component = ['Cluster2'])
RF_NN_Add_cluster25 = Experiment_NN(xtrain = RF_Credit_card_train_ADD_Cluter2_5_10_25, ytrain = RF_CreditCard_trainY,
xtest=RF_Credit_card_Test_ADD_Cluter2_5_10_25, ytest=RF_CreditCard_TestY,
cluster_feature=[], num_of_component = ['Cluster25'])
print('PCA_NN_Add_cluster5_25',PCA_NN_Add_cluster25)
print('ICA_NN_Add_cluster2_10',ICA_NN_Add_cluster10)
print('RP_NN_Add_cluster2_10',RP_NN_Add_cluster2)
print('RF_NN_Add_cluster10_25',RF_NN_Add_cluster25)