-
Notifications
You must be signed in to change notification settings - Fork 2
/
lvq3.py
317 lines (274 loc) · 11.7 KB
/
lvq3.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
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
# -*- coding: utf-8 -*-
"""
Created on Fri Sep 21 22:05:22 2018
@author: USER
"""
# IMPORT LIBRARY
import numpy as np
import csv
import time
from datetime import datetime
import os
#dataClass = [[[1.0,1.0,0.0,0.0],0],[[0.0,0.0,0.0,1.0],1]]
#dataTraining = [[[0.0,0.0,1.0,1.0],1],[[1.0,0.0,0.0,0.0],0],[[0.0,1.0,1.0,0.0],1]]
# SET PARAMETER VALUES
alpha_init = 0.2
alpha = alpha_init
pengali_alpha = 0.8
m = 0.1
beta = m*alpha
#epsilon= 0.35
epsilon2= 0.4
start_time = time.time()
iterasi_maksimal = 10
alpha_minimal = 0.000001
iterasi = 1
#epsilon2= 0.8
# OPEN CSV FILE (INPUT)
def read_csv(file_name):
array_2D = []
with open(file_name, 'rb') as csvfile:
read = csv.reader(csvfile, delimiter=';')
for row in read:
array_2D.append(row)
return array_2D
# SPLIT DATA INTO TRAINING DATA AND TESTING DATA
def buildFold(foldTest):
path = "dataBaru/"
#pathOutput = "D:\\KULIAH\\SEMESTER VII\\SKRIPSI - OFFLINE\\DATA01\\"
tot = os.listdir(path)
#kelas = 1
fold = []
for filename in tot:
fold.append(read_csv(path+filename)) # Fold 1
#foldTest = [0]
dataUji = []
for i in range(len(foldTest)):
dataUji.extend(fold[foldTest[i]])
dataLatih = []
for i in range(len(fold)):
if (i not in foldTest):
dataLatih.extend(fold[i])
return dataLatih, dataUji
# DO FEATURE SELECTION FROM BOTH TRAINING AND TESTING DATA
def featureSelection(listFeatures):
res = np.transpose(listFeatures)
result = []
#eliminated = [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] # Warna
#eliminated = [41,42,43,44,45,46,47,48,49,50,51,52] # Haralick
#eliminated = [4,5,14,15,24,25,26,34,35] # Relief
#eliminated = [1,2,9,10,11,12,19,20,21,22,29,30,31,32,39,40,41,42,44,45,46,48,49] # Korelasi
#eliminated = [1,2,3,5,6,8,9,10,11,13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37,38,39,40,43,46,47,48,49,50,51,52] # CFS
eliminated = [] # Tanpa Seleksi
for i in range(len(res)):
if ((i+1) not in eliminated):
resJ = []
for j in range(len(res[i])):
resJ.append(res[i][j])
result.append(resJ)
return np.transpose(result).tolist()
# DO PCA FEATURE SELECTION FROM BOTH TRAINING AND TESTING DATA
def getPCAFeatures(listFeatures):
res = listFeatures
result = []
for i in range(len(res)):
fiturPCA = []
fiturPCA.append((0.209*res[i][13])+(0.209*res[i][12])+(0.208*res[i][32])+(0.208*res[i][33])+(0.208*res[i][17]))
fiturPCA.append((-0.293*res[i][11])-(0.293*res[i][10])-(0.293*res[i][30])-(0.293*res[i][31])-(0.292*res[i][21]))
fiturPCA.append((-0.301*res[i][5])-(0.301*res[i][4])-(0.299*res[i][25])-(0.299*res[i][24])-(0.297*res[i][34]))
fiturPCA.append((-0.386*res[i][47])-(0.385*res[i][45])-(0.352*res[i][49])-(0.352*res[i][43])-(0.345*res[i][51]))
fiturPCA.append((0.348*res[i][18])+(0.345*res[i][28])+(0.344*res[i][8])+(0.341*res[i][38])-(0.264*res[i][40]))
fiturPCA.append((0.391*res[i][39])+(0.37*res[i][19])+(0.369*res[i][9])+(0.366*res[i][29])-(0.283*res[i][8]))
fiturPCA.append((-0.41*res[i][41])-(0.393*res[i][46])+(0.373*res[i][51])-(0.349*res[i][43])-(0.34*res[i][40]))
fiturPCA.append((0.607*res[i][40])-(0.53*res[i][42])+(0.22*res[i][45])+(0.212*res[i][47])-(0.208*res[i][43]))
result.append(fiturPCA)
return result
#data1 = read_csv('data/realTraining/dataTrain80204.csv') # Data Training
#data3 = read_csv('data/realTraining/dataTest80204.csv') # Data Testing
# SET REFERENCE VECTOR, TRAINING DATA AND TESTING DATA
data2 = read_csv('data/dataClass.csv') # Data Class (Vector Reference)
testFold = [0]
fold = 'fold0'
# PRINT PARAMETER VALUES
result = []
result.append('fold='+str(fold))
result.append('alpha init = '+str(alpha_init))
result.append('pengali_alpha = '+str(pengali_alpha))
result.append('m = '+str(m))
result.append('epsilon = '+str(epsilon2))
result.append('iterasi_maksimal = '+str(iterasi_maksimal))
result.append('alpha_minimal = '+str(alpha_minimal))
result.append('fitur = all')
#result.append('fitur = tekstur')
#result.append('fitur = warna')
#result.append('waktu = '+str(waktu))
for i in result:
print(i)
# SPLIT AND SET TRAINING DATA AND TESTING DATA
data1, data3 = buildFold(testFold)
'''
data1 = read_csv('data/realTraining/dataTrain9010.csv') # Data Training
data3 = read_csv('data/realTraining/dataTest9010.csv') # Data Testing
data2 = read_csv('data/realTraining/dataClass80204.csv') # Data Class (Vector Reference)
'''
'''
data1 = read_csv('data/pengujian124/dataTrain124.csv') # Data Training
data2 = read_csv('data/pengujian124/dataClass124.csv') # Data Class (Vector Reference)
data3 = read_csv('data/pengujian124/dataTest124.csv') # Data Testing
'''
'''
data1 = read_csv('data/dataTrainAll50.csv') # Data Training
data2 = read_csv('data/dataClassAllTL.csv') # Data Class (Vector Reference)
data3 = read_csv('data/dataTestAll50.csv') # Data Testing
data1 = read_csv('datatraining.csv') # Data Training
data2 = read_csv('refvector.csv') # Data Class (Vector Reference)
data3 = read_csv('datatesting.csv') # Data Testing
'''
# SPLIT DATA INTO DEPENDENT VARIABLE (CLASS) AND INDEPENDENT VARIABLE (FEATURES)
dataTrain = ((np.array(data1[:]))[:,1:-1]).astype(np.float64).tolist()
dataC = ((np.array(data2[:]))[:,1:-1]).astype(np.float64).tolist()
dataT = ((np.array(data3[:]))[:,1:-1]).astype(np.float64).tolist()
classDataTrain = ((np.array(data1[:]))[:,-1:]).astype(int).tolist()
classDataClass = ((np.array(data2[:]))[:,-1:]).astype(int).tolist()
classDataTest = ((np.array(data3[:]))[:,-1:]).astype(int).tolist()
dataTraining = []
dataClass = []
'''
dataC = getPCAFeatures(dataC)
dataTrain = getPCAFeatures(dataTrain)
dataT = getPCAFeatures(dataT)
'''
# DO FEATURE SELECTION
dataC = featureSelection(dataC)
dataTrain = featureSelection(dataTrain)
dataT = featureSelection(dataT)
dataTesting = []
ignoredClass = [] # eliminated Class
#ignoredClass = [8,22,23,25,27,29]
for i in range(len(dataTrain)):
if (classDataTrain[i][0] not in ignoredClass):
dataArray = []
dataArray.append(dataTrain[i])
dataArray.append(classDataTrain[i][0])
dataTraining.append(dataArray)
for i in range(len(dataC)):
if (classDataClass[i][0] not in ignoredClass):
dataArray2 = []
dataArray2.append(dataC[i])
dataArray2.append(classDataClass[i][0])
dataClass.append(dataArray2)
#dataTesting.append(dataT[i])
for i in range(len(dataT)):
dataTesting.append(dataT[i])
# Initialize Weight Matrix
weightMatrix = np.zeros((len(dataClass),len(dataTraining[0][0])),dtype=np.float64)
for i in range(len(weightMatrix)):
for j in range(len(dataTraining[i][0])):
weightMatrix[i][j] = dataClass[i][0][j]
def takeFirst(elem):
return elem[0]
countLVQ1 = 0
countLVQ2 = 0
countLVQ21 = 0
countLVQ3 = 0
# ITERATION LVQ
#for x in range(iterasi):
while(iterasi <= iterasi_maksimal and alpha >= alpha_minimal):
# Find Euclidean Distance
jarak = np.zeros(len(dataClass),dtype=np.float64)
for i in range(len(dataTraining)):
for j in range(len(dataClass)):
jarak[j] = 0
for k in range(len(dataTraining[i][0])):
jarak[j] += np.power(dataTraining[i][0][k] - weightMatrix[j][k],2)
jarak[j] = np.sqrt(jarak[j])
jVal = []
for j in range(len(dataClass)):
jVal.append([jarak[j],j])
jVal.sort(key=takeFirst)
t = dataTraining[i][1]
yc1 = int(jVal[0][1]) # The nearest class from the data training
yc2 = int(jVal[1][1]) # The second nearest class from the data training
dc1 = jVal[0][0] # The distance of the nearest class from the data training
dc2 = jVal[1][0] # The distance of the second nearest class from the data training
for j in range(len(weightMatrix)):
for k in range(len(weightMatrix[j])):
if ( (min(np.divide(dc1,dc2),np.divide(dc2,dc1)) > ((1-epsilon2)*(1+epsilon2)))):
if ((classDataClass[yc1][0] == t and classDataClass[yc2][0] == t) ):
#print(str(t)+" "+str(yc1)+" "+str(yc2))
countLVQ3 += 1
if(j == yc1 or j == yc2):
weightMatrix[j][k] = (1-beta)*weightMatrix[j][k] + beta*dataTraining[i][0][k]
elif((classDataClass[yc1][0] == t and classDataClass[yc2][0] != t) or (classDataClass[yc1][0] != t and classDataClass[yc2][0] == t)):
countLVQ21 += 1
if(j == yc1):
if(t == classDataClass[yc1][0]):
weightMatrix[j][k] = (1-alpha)*weightMatrix[j][k] + alpha*dataTraining[i][0][k]
else:
weightMatrix[j][k] = (1+alpha)*weightMatrix[j][k] - alpha*dataTraining[i][0][k]
elif(j == yc2):
if(t == classDataClass[yc2][0]):
weightMatrix[j][k] = (1-alpha)*weightMatrix[j][k] + alpha*dataTraining[i][0][k]
else:
weightMatrix[j][k] = (1+alpha)*weightMatrix[j][k] - alpha*dataTraining[i][0][k]
'''
elif (j == yc1):
countLVQ1 += 1
if(classDataClass[j][0] == t):
weightMatrix[j][k] = (1-alpha)*weightMatrix[j][k] + alpha*dataTraining[i][0][k]
else:
weightMatrix[j][k] = (1+alpha)*weightMatrix[j][k] - alpha*dataTraining[i][0][k]
'''
alpha = pengali_alpha*alpha
iterasi += 1
with open('data/weight.csv', 'a') as myfile:
wr = csv.writer(myfile, delimiter=',')
for i in range(len(weightMatrix)):
result = weightMatrix[i]
wr.writerow(result)
# Testing
#dataTest = [0.0,0.0,1.0,1.0]
wrongClass = 0
#with open('data/realTesting/testingResultMValueAll04.csv', 'a') as myfile:
with open('data/'+fold+'Testing.csv', 'a') as myfile:
wr = csv.writer(myfile, delimiter=',')
for z in range(len(dataTesting)):
testing = classDataTest[z][0]-1
dataTest = dataTesting[testing]
classResult = -1
minValue = 999999999
for i in range(len(weightMatrix)):
sumValue = 0
for j in range(len(weightMatrix[i])):
sumValue += np.power(dataTesting[z][j] - weightMatrix[i][j],2)
if (sumValue < minValue):
minValue = sumValue
#classResult = i
classResult = classDataClass[i][0]-1
result = [testing,classResult]
wr.writerow(result)
if (testing != classResult):
#print('Real Class = '+str(testing)+' -> Class Result = '+str(classResult)+' -> Min Value = '+str(minValue))
wrongClass+=1
akurasi = np.divide(float(len(dataTesting) - wrongClass),float(len(dataTesting)))
waktu = time.time()-start_time
print akurasi
'''
with open('data/crossValidation/akurasi/RESULT.csv', 'a') as myfile:
wr = csv.writer(myfile, delimiter=',')
result = [fold,akurasi]
wr.writerow(result)
#print(countLVQ1)
#print(countLVQ2)
#print(countLVQ21)
#print(countLVQ3)
with open('data/crossValidation/akurasi/TIME_COMPUTATION.csv', 'a') as myfile:
wr = csv.writer(myfile, delimiter=',')
date = datetime.now()
result.append(date)
result.append('alpha akhir = '+str(alpha))
result.append('iterasi = '+str(iterasi))
print('alpha akhir = '+str(alpha))
print('iterasi = '+str(iterasi))
wr.writerow(result)
'''