/
experiment.py
284 lines (259 loc) · 13.2 KB
/
experiment.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
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
#define streaming experiments
import parameters
import classifier as c
import preprocess
import miscMethods as misc
import preprocess2
import random
import copy
import olsf
import numpy as np
import matplotlib.pyplot as plt
import time
#Non Changable Parameters
lambda_error=0.5#weight of hinge loss, 1-weight of regularization
lambda_confidence=1-lambda_error
parameter_r_error=1/2*lambda_error
parameter_r_confidence=1/2*lambda_confidence
#sparsity
random.seed(parameters.seed)
param_lambda=30
C=1
B=1
def parameterSearch():
global B
global lambda_confidence
global parameter_r_error
global parameter_r_confidence
global lambda_error
lambda_error=0.99
for i in range (0, 10):
B=0
for j in range (0, 10):
compareClassifiersOverUCIDatasets()
B+=0.1
lambda_confidence=1-lambda_error
parameter_r_error=1/2*lambda_error
parameter_r_confidence=1/2*lambda_confidence
lambda_error-=0.1
def compareClassifiersOverUCIDatasets():
compare_classifiers(preprocess2.readGermanNormalized(), "German")
#compare_classifiers(preprocess2.readIonosphereNormalized(), "Ionosphere")
#compare_classifiers(preprocess2.readMagicNormalized(), "Magic")
#compare_classifiers(preprocess2.readSpambaseNormalized(), "Spambase")
#compare_classifiers(preprocess2.readWdbcNormalized(), "WDBC")
#compare_classifiers(preprocess2.readWpbcNormalized(), "WPBC")
#compare_classifiers(preprocess2.readA8ANormalized(), "A8A")
#compare_classifiers(preprocess2.readSvmguide3Normalized(), "svmguide3")
def compare_classifiers(readed_dataset, dataset_name):
plot_list=[("OLVF_random", variableFeatureExperiment, parameters.olvf_random),
("OLVF_trapez", trapezoidalExperiment, parameters.olvf_trapez),
("OLVF_trapez_shuffled_sparse", trapezoidalExperimentShuffledSparse, parameters.olvf_trapez_shuffled),
("OLSF", trapezoidalExperimentOLSF, parameters.olsf),
("OLVF_random_sparse", variableFeatureExperiment, parameters.olvf_random_sparse),
("OLVF_trapez_sparse", trapezoidalExperimentSparse, parameters.olvf_trapez_sparse)]
x = list(range(len(readed_dataset)))
plotlist=[]
for triple in plot_list:
if triple[2]==1:
plotlist.append((triple[1](readed_dataset, dataset_name), triple[0]))
for i in range(len(plotlist)):
plt.plot(x[2:], plotlist[i][0][2:], label=plotlist[i][1])
plt.legend()
plt.xlabel("Instance")
plt.ylabel("Test Error Rate %")
plt.title(dataset_name)
plt.savefig('./figures/'+'compare_'+str(lambda_error)+str("_")+str(B)+str("_")+time.strftime("%H%M%S")+'.png')
plt.show()
def variableFeatureExperiment(input_dataset, dataset_name, mode="OLVF_random_sparse"):
print("Variable feature experiment with OLVF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeRandomData(copy.deepcopy(input_dataset))]
current_dataset= preprocess2.removeRandomData(copy.deepcopy(input_dataset))
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(current_dataset)
current_classifier=c.classifier(current_dataset, [])
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.mean_dict)
return average_error_vector
def variableFeatureExperimentSparse(input_dataset, dataset_name):
print("Variable feature experiment with OLVF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeRandomData(copy.deepcopy(input_dataset))]
current_dataset= preprocess2.removeRandomData(copy.deepcopy(input_dataset))
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(current_dataset)
current_classifier=c.classifier(current_dataset, [], 1)
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.mean_dict)
return average_error_vector
def trapezoidalExperiment(input_dataset, dataset_name):
print("Trapezoidal experiment with OLVF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))]
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(input_dataset)
current_dataset= preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))
current_classifier= c.classifier(current_dataset, [])
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.mean_dict)
return average_error_vector
def trapezoidalExperimentSparse(input_dataset, dataset_name):
print("Trapezoidal experiment with OLVF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))]
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(input_dataset)
current_dataset= preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))
current_classifier= c.classifier(current_dataset, [], 1)
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.mean_dict)
return average_error_vector
def trapezoidalExperimentShuffled(input_dataset, dataset_name):
print("Trapezoidal experiment shuffled with OLVF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))]
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(input_dataset)
current_dataset= preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))
random.shuffle(current_dataset)
current_classifier= c.classifier(current_dataset, [])
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.mean_dict)
return average_error_vector
def trapezoidalExperimentShuffledSparse(input_dataset, dataset_name):
print("Trapezoidal experiment shuffled with OLVF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))]
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(input_dataset)
current_dataset= preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))
random.shuffle(current_dataset)
current_classifier= c.classifier(current_dataset, [], 1)
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.mean_dict)
return average_error_vector
def trapezoidalExperimentOLSF(input_dataset, dataset_name):
print("Trapezoidal experiment with OLSF: "+str(dataset_name))
error_vector=np.zeros(len(input_dataset))
feature_summary=[len(row) for row in preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))]
for i in range(parameters.rounds):
print("Round: "+str(i))
random.seed(parameters.seed)
random.shuffle(input_dataset)
current_dataset= preprocess2.removeDataTrapezoidal(copy.deepcopy(input_dataset))
current_classifier= olsf.classifier(current_dataset, [])
classifier_summary, stream_error= current_classifier.train()
error_vector=np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
#misc.plotError(average_error_vector, dataset_name)
#misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
print(current_classifier.weight_dict)
return average_error_vector
def Enron_experiment():
print("Enron Experiment")
for index, folder_tuple in enumerate(parameters.enron_folders): #iterate over all Enron folders
dataset=(preprocess.data(folder_tuple[0], folder_tuple[1])).dataset
error_vector=np.zeros(len(dataset))
feature_summary=[len(i) for i in dataset] #extract the feature dimension vector to plot
dataset_name="Enron"+str(index+1)
dataset= olvf_closed_form.preprocessData(dataset)
my_classifier=olvf_closed_form.olvf("stream")
classifier_summary, stream_error =my_classifier.train()
error_vector= np.add(error_vector, stream_error)
misc.plotError(error_vector, dataset_name)
misc.plotFeatures(feature_summary, dataset_name)
#misc.plotClassifierDimension(classifier_summary, dataset_name)
#print(my_classifier.mean_dict)
#return average_error_vector
def url_experiment_olsf():
print("URL Dataset Experiment")
dataset_name= "URL"
error_vector=[]
weights= {}
for day in range(0, 121):
print("File: Day"+str(day))
dataset=preprocess2.readUrlNormalized(day)
my_classifier=olsf.classifier(dataset, [])
my_classifier.weight_dict= weights
classifier_summary, stream_error =my_classifier.train()
weights=my_classifier.weight_dict
error_vector= np.append(error_vector,stream_error)
misc.plotError(error_vector[0::2500], dataset_name)
return error_vector
def url_experiment_olvf():
print("URL Dataset Experiment")
dataset_name= "URL"
error_vector=[]
weights= {}
for day in range(0, 121):
print("File: Day"+str(day))
dataset=preprocess2.readUrlNormalized(day)
my_classifier=c.classifier(dataset, [], 1)
my_classifier.weight_dict= weights
classifier_summary, stream_error =my_classifier.train()
weights=my_classifier.mean_dict
error_vector= np.append(error_vector,stream_error)
misc.plotError(error_vector[0::2500], dataset_name)
return error_vector
def TREC_experiment():
print("TREC Experiment")
dataset=(preprocess.data(parameters.spam_email[0], parameters.ham_email[0])).dataset
error_vector=np.zeros(len(dataset))
feature_summary=[len(i) for i in dataset] #extract the feature dimension vector to plot
dataset_name="TREC"
for i in range(0, parameters.rounds):
print("Round: "+str(i))
random.shuffle(dataset) #shuffle and split dataset
my_classifier=c.classifier(dataset, [], 1)
classifier_summary, stream_error =my_classifier.train()
error_vector= np.add(error_vector, stream_error)
average_error_vector= np.divide(error_vector, parameters.rounds)
misc.plotError(average_error_vector[1:], dataset_name)
misc.plotFeatures(feature_summary, dataset_name)
misc.plotClassifierDimension(classifier_summary, dataset_name)
#print(my_classifier.mean_dict)
#return average_error_vector