Neural Networks

Kategori: Machine Learning , 21 Aralık 2019 , JanFranco


Yapay sinir ağları algoritması, insan beyninin basit bir simülasyonu gibi düşünülebilir. Bu yazımda Keras kütüphanesini kullanarak Python üzerinde bir yapay sinir ağı modeli oluşturacağız ve train edeceğiz. Yapay sinir ağları ile detaylı bilgiyi Deep Learning bölümünde anlattım. Bu nedenle tekrar üstünde durmuyorum. İlk olarak kütüphaneleri ve veriyi import edelim:


import numpy as np
import pandas as pd
from keras.layers import Dense
from keras.models import Sequential
from sklearn.metrics import confusion_matrix
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import LabelEncoder, OneHotEncoder, StandardScaler

pd.set_option('display.max_columns', 500)
pd.set_option('display.width', 1000)

df = pd.read_csv('Churn_Modelling.csv')
print(df)

>>

      RowNumber  CustomerId         Surname  CreditScore Geography  Gender  Age  Tenure    Balance  NumOfProducts  HasCrCard  IsActiveMember  EstimatedSalary  Exited
0             1    15634602        Hargrave          619    France  Female   42       2       0.00              1          1               1        101348.88       1
1             2    15647311            Hill          608     Spain  Female   41       1   83807.86              1          0               1        112542.58       0
2             3    15619304            Onio          502    France  Female   42       8  159660.80              3          1               0        113931.57       1
3             4    15701354            Boni          699    France  Female   39       1       0.00              2          0               0         93826.63       0
4             5    15737888        Mitchell          850     Spain  Female   43       2  125510.82              1          1               1         79084.10       0
5             6    15574012             Chu          645     Spain    Male   44       8  113755.78              2          1               0        149756.71       1
6             7    15592531        Bartlett          822    France    Male   50       7       0.00              2          1               1         10062.80       0
7             8    15656148          Obinna          376   Germany  Female   29       4  115046.74              4          1               0        119346.88       1
8             9    15792365              He          501    France    Male   44       4  142051.07              2          0               1         74940.50       0
9            10    15592389              H?          684    France    Male   27       2  134603.88              1          1               1         71725.73       0
10           11    15767821          Bearce          528    France    Male   31       6  102016.72              2          0               0         80181.12       0
11           12    15737173         Andrews          497     Spain    Male   24       3       0.00              2          1               0         76390.01       0
12           13    15632264             Kay          476    France  Female   34      10       0.00              2          1               0         26260.98       0
13           14    15691483            Chin          549    France  Female   25       5       0.00              2          0               0        190857.79       0
14           15    15600882           Scott          635     Spain  Female   35       7       0.00              2          1               1         65951.65       0
15           16    15643966         Goforth          616   Germany    Male   45       3  143129.41              2          0               1         64327.26       0
16           17    15737452           Romeo          653   Germany    Male   58       1  132602.88              1          1               0          5097.67       1
17           18    15788218       Henderson          549     Spain  Female   24       9       0.00              2          1               1         14406.41       0
18           19    15661507         Muldrow          587     Spain    Male   45       6       0.00              1          0               0        158684.81       0
19           20    15568982             Hao          726    France  Female   24       6       0.00              2          1               1         54724.03       0
20           21    15577657        McDonald          732    France    Male   41       8       0.00              2          1               1        170886.17       0
21           22    15597945        Dellucci          636     Spain  Female   32       8       0.00              2          1               0        138555.46       0
22           23    15699309       Gerasimov          510     Spain  Female   38       4       0.00              1          1               0        118913.53       1
23           24    15725737          Mosman          669    France    Male   46       3       0.00              2          0               1          8487.75       0
24           25    15625047             Yen          846    France  Female   38       5       0.00              1          1               1        187616.16       0
25           26    15738191         Maclean          577    France    Male   25       3       0.00              2          0               1        124508.29       0
26           27    15736816           Young          756   Germany    Male   36       2  136815.64              1          1               1        170041.95       0
27           28    15700772         Nebechi          571    France    Male   44       9       0.00              2          0               0         38433.35       0
28           29    15728693      McWilliams          574   Germany  Female   43       3  141349.43              1          1               1        100187.43       0
29           30    15656300        Lucciano          411    France    Male   29       0   59697.17              2          1               1         53483.21       0
...         ...         ...             ...          ...       ...     ...  ...     ...        ...            ...        ...             ...              ...     ...
9970       9971    15587133        Thompson          518    France    Male   42       7  151027.05              2          1               0        119377.36       0
9971       9972    15721377            Chou          833    France  Female   34       3  144751.81              1          0               0        166472.81       0
9972       9973    15747927           Ch'in          758    France    Male   26       4  155739.76              1          1               0        171552.02       0
9973       9974    15806455          Miller          611    France    Male   27       7       0.00              2          1               1        157474.10       0
9974       9975    15695474          Barker          583    France    Male   33       7  122531.86              1          1               0         13549.24       0
9975       9976    15666295           Smith          610   Germany    Male   50       1  113957.01              2          1               0        196526.55       1
9976       9977    15656062         Azikiwe          637    France  Female   33       7  103377.81              1          1               0         84419.78       0
9977       9978    15579969         Mancini          683    France  Female   32       9       0.00              2          1               1         24991.92       0
9978       9979    15703563           P'eng          774    France    Male   40       9   93017.47              2          1               0        191608.97       0
9979       9980    15692664          Diribe          677    France  Female   58       1   90022.85              1          0               1          2988.28       0
9980       9981    15719276            T'ao          741     Spain    Male   35       6   74371.49              1          0               0         99595.67       0
9981       9982    15672754        Burbidge          498   Germany    Male   42       3  152039.70              1          1               1         53445.17       1
9982       9983    15768163         Griffin          655   Germany  Female   46       7  137145.12              1          1               0        115146.40       1
9983       9984    15656710           Cocci          613    France    Male   40       4       0.00              1          0               0        151325.24       0
9984       9985    15696175  Echezonachukwu          602   Germany    Male   35       7   90602.42              2          1               1         51695.41       0
9985       9986    15586914          Nepean          659    France    Male   36       6  123841.49              2          1               0         96833.00       0
9986       9987    15581736        Bartlett          673   Germany    Male   47       1  183579.54              2          0               1         34047.54       0
9987       9988    15588839         Mancini          606     Spain    Male   30       8  180307.73              2          1               1          1914.41       0
9988       9989    15589329         Pirozzi          775    France    Male   30       4       0.00              2          1               0         49337.84       0
9989       9990    15605622        McMillan          841     Spain    Male   28       4       0.00              2          1               1        179436.60       0
9990       9991    15798964      Nkemakonam          714   Germany    Male   33       3   35016.60              1          1               0         53667.08       0
9991       9992    15769959     Ajuluchukwu          597    France  Female   53       4   88381.21              1          1               0         69384.71       1
9992       9993    15657105     Chukwualuka          726     Spain    Male   36       2       0.00              1          1               0        195192.40       0
9993       9994    15569266          Rahman          644    France    Male   28       7  155060.41              1          1               0         29179.52       0
9994       9995    15719294            Wood          800    France  Female   29       2       0.00              2          0               0        167773.55       0
9995       9996    15606229        Obijiaku          771    France    Male   39       5       0.00              2          1               0         96270.64       0
9996       9997    15569892       Johnstone          516    France    Male   35      10   57369.61              1          1               1        101699.77       0
9997       9998    15584532             Liu          709    France  Female   36       7       0.00              1          0               1         42085.58       1
9998       9999    15682355       Sabbatini          772   Germany    Male   42       3   75075.31              2          1               0         92888.52       1
9999      10000    15628319          Walker          792    France  Female   28       4  130142.79              1          1               0         38190.78       0
pd.set_option() methodu ile pandas kütüphanesinin konsol ayarlarını değiştirdik. Böylelikle yukarıdaki çıktıyı alabildik. Eğer bu ayarları kullanmasaydık bir çok sütun konsola basılmayacaktı. Burada gördüğümüz veriler, müşteri analizi verileridir. Bir müşterinin numarası, soyadı, kredi skoru, ülkesi, cinsiyeti, yaşı, kaç yıldır müşteri olduğu, bakiyesi, ürünleri, kredi kartı olup olmadığı, maaşı gibi bilgiler mevcut. Son sütunda ise müşterinin bankayı bırakıp bırakmamasının verileri mevcut. Ön işlemede One Hot Encoding, Label Encoding ve Scaling işlemleri yapacağız. Geograpy sütununa One Hot Encoding işlemini, Gender sütununa Label Encoding işlemini, CreditScore, Age, Tenure, Balance, NumOfProducts, EstimatedSalary sütunlarına da Scaling işlemini yapacağız. Tahmin edeceğimiz sütun Exited sütunu olacak. Bu saydığım sütunlara ek olarak HasCrCard ve IsActiveMember sütunlarını alacağız. Kalan sütunlarla bir işimiz yok. Preprocessing işlemlerini bir çok kez anlattığım için direk kodu veriyorum, anlatmadan geçiyorum:


le = LabelEncoder()
ohe = OneHotEncoder()
df.iloc[:, [5]] = le.fit_transform(df.iloc[:, [5]])
geography = df.iloc[:, [4]].values
geography = ohe.fit_transform(geography).toarray()
result = pd.DataFrame(data=geography, index=range(10000), columns=['fr', 'sp', 'ge'])

sc = StandardScaler()
mustScale = df.iloc[:, [3, 6, 7, 8, 9, 12]].values
mustScale = sc.fit_transform(mustScale)
result2 = pd.DataFrame(data=mustScale, index=range(10000), columns=['creditScore', 'age', 'tenure', 'balance',
                                                                    'numOfProducts', 'estimatedSalary'])

CardActive = df.iloc[:, [10, 11]]
x = pd.concat([result, result2, CardActive], axis=1)
y = df.iloc[:, [13]].values

x_train, x_test, y_train, y_test = train_test_split(x, y, test_size=.33, random_state=0)
print(x_train)
print(y_train)

>>

       fr   sp   ge  creditScore       age    tenure   balance  numOfProducts  estimatedSalary  HasCrCard  IsActiveMember
6596  1.0  0.0  0.0    -1.795464 -1.613554 -0.004426  0.097341       0.807737         1.095901          0               0
1292  1.0  0.0  0.0     1.484464  0.198164  0.687130 -1.225848       0.807737        -0.448900          1               1
2011  1.0  0.0  0.0    -0.833213 -0.660018  1.378686  0.654734       0.807737        -0.648250          0               1
4957  1.0  0.0  0.0     1.049900  0.007457 -0.695982 -1.225848      -0.911583         1.531915          1               0
5164  1.0  0.0  0.0    -0.088246 -0.278604 -0.350204 -1.225848       0.807737         1.654297          1               1
5655  0.0  0.0  1.0     0.460134  2.295943 -1.387538  1.117363      -0.911583         0.462523          1               1
3617  0.0  1.0  0.0     1.463771 -0.946079 -0.350204  0.498698      -0.911583        -1.411431          1               0
9058  0.0  0.0  1.0    -1.867891  0.865639  0.341352  1.027181      -0.911583         0.293540          1               1
5971  0.0  0.0  1.0     1.598279 -0.183251 -0.004426 -1.225848       0.807737        -1.359149          1               0
1953  1.0  0.0  0.0    -0.284834 -1.708908  1.724464 -1.225848       0.807737         0.621849          0               1
9158  0.0  0.0  1.0    -1.143617 -0.755372  1.724464  0.667895      -0.911583         1.007472          1               1
9246  1.0  0.0  0.0     1.153368 -1.804262 -1.387538  1.008853      -0.911583        -0.251852          1               1
3000  1.0  0.0  0.0    -1.226391  0.102810 -0.004426 -1.225848       0.807737         1.339118          0               1
1343  0.0  0.0  1.0     0.201464  0.293517 -0.695982  0.081789      -0.911583         1.532624          1               0
917   1.0  0.0  0.0    -0.046858  0.579578 -0.695982 -0.470414      -0.911583        -0.745185          1               1
5175  0.0  0.0  1.0     0.501521 -0.946079  1.378686  0.818746       0.807737         1.124487          1               0
5396  1.0  0.0  0.0    -0.615931  0.293517  1.378686  1.364956       0.807737         0.550601          1               1
4635  1.0  0.0  0.0    -0.326221 -0.564665 -1.041760  1.462436       0.807737        -0.606316          1               1
1336  0.0  0.0  1.0    -0.336568 -0.087897 -0.004426  0.801165      -0.911583        -0.141160          1               0
8951  0.0  1.0  0.0     0.532561 -1.518201  0.687130  0.357234      -0.911583         0.252379          1               0
4732  0.0  1.0  0.0    -0.129633  1.342407 -1.387538  0.760178      -0.911583        -1.450809          1               0
4526  1.0  0.0  0.0     1.525851 -0.469311  1.378686  1.250273      -0.911583         1.587714          1               0
2586  1.0  0.0  0.0     0.418746  1.056346  0.341352  0.969120      -0.911583        -1.693432          1               1
9507  1.0  0.0  0.0     1.629319  0.198164 -1.733315 -1.225848      -0.911583        -0.351283          1               1
5229  1.0  0.0  0.0     0.253198 -0.373958  1.032908 -1.225848       0.807737        -0.755505          1               1
3669  1.0  0.0  0.0     0.025569 -0.946079 -1.041760 -1.225848       0.807737        -1.015838          1               1
85    0.0  0.0  1.0     0.015222  3.440186  1.724464 -1.225848       0.807737         0.253627          1               1
3319  1.0  0.0  0.0    -2.281762 -0.373958 -1.387538  0.679683      -0.911583        -1.689631          0               0
9665  1.0  0.0  0.0    -1.071189 -0.850726  0.341352 -1.225848       0.807737        -1.419262          1               1
5032  0.0  1.0  0.0     0.346319  0.674932 -0.695982  0.424226       0.807737         0.939825          1               0
...   ...  ...  ...          ...       ...       ...       ...            ...              ...        ...             ...
755   1.0  0.0  0.0     1.577585 -0.755372  1.378686  1.292345       0.807737         0.716606          1               0
8291  0.0  0.0  1.0     0.180771 -0.087897 -1.387538  1.144631      -0.911583        -0.534193          1               1
2496  1.0  0.0  0.0    -0.864254  0.579578 -0.350204 -1.225848       0.807737         0.364525          0               1
7599  1.0  0.0  0.0    -1.174657  0.579578  0.687130  1.316389      -0.911583         0.361739          1               0
1871  0.0  0.0  1.0    -0.171020 -0.278604 -1.041760  0.152423      -0.911583        -1.464198          1               0
2046  0.0  1.0  0.0    -1.350552 -0.469311  1.032908  0.697708       0.807737        -1.734498          1               1
7877  0.0  0.0  1.0    -1.619568  0.198164 -1.041760 -0.104356       0.807737         1.536073          1               0
4851  1.0  0.0  0.0    -1.247084  0.865639 -1.733315 -1.225848      -0.911583        -0.260994          1               0
5072  1.0  0.0  0.0    -1.847197 -0.946079  1.032908  0.416775      -0.911583         1.410845          0               1
2163  1.0  0.0  0.0    -0.471076 -0.373958  0.341352 -1.225848       0.807737        -0.954372          1               1
6036  1.0  0.0  0.0     0.242851 -1.136786 -0.350204  0.042607      -0.911583        -0.858085          0               1
6921  0.0  0.0  1.0     0.294585  0.007457  0.687130  0.235631      -0.911583        -1.634825          0               1
6216  1.0  0.0  0.0    -1.774770  0.770285 -1.387538 -1.225848      -0.911583        -0.083805          1               0
537   0.0  0.0  1.0     0.367013 -0.469311  1.378686 -1.225848       0.807737         0.477842          1               0
9893  1.0  0.0  0.0     1.174061  0.102810  1.378686  0.384565      -0.911583         0.417431          1               0
2897  0.0  1.0  0.0    -1.267778 -0.946079  1.378686  1.502194       0.807737        -0.883470          1               0
7768  0.0  0.0  1.0    -0.512463 -1.804262  1.378686  0.736617       0.807737        -0.231245          1               0
2222  0.0  1.0  0.0     0.625682  0.388871 -1.041760 -0.600097       0.807737         1.309973          1               1
2599  0.0  0.0  1.0    -1.040149  2.200589  0.687130  0.071187      -0.911583        -0.619041          0               1
705   0.0  1.0  0.0     1.029206  0.102810 -0.004426  1.471303       2.527057        -1.384830          1               0
3468  1.0  0.0  0.0     1.753480  0.960993 -1.387538 -1.225848       0.807737         0.330339          1               1
6744  1.0  0.0  0.0     1.308569 -0.850726 -0.350204 -1.225848       0.807737         0.269907          0               1
5874  1.0  0.0  0.0    -0.553850 -1.899615 -1.041760 -1.225848       0.807737        -0.157431          1               1
4373  0.0  0.0  1.0    -1.081536  0.579578 -1.041760 -1.225848      -0.911583         1.698898          0               0
7891  0.0  1.0  0.0     0.284238  0.865639 -1.041760  0.394491       0.807737         1.623886          0               1
9225  0.0  1.0  0.0    -0.584891 -0.660018 -0.350204  0.698607       0.807737         1.093273          1               1
4859  0.0  0.0  1.0     1.484464 -1.613554 -0.350204  0.608299      -0.911583         0.133249          1               1
3264  1.0  0.0  0.0     0.905045 -0.373958 -0.004426  1.358909       0.807737         1.414415          1               0
9845  0.0  0.0  1.0    -0.626278 -0.087897  1.378686 -1.225848       0.807737         0.846147          1               1
2732  0.0  1.0  0.0    -0.284834  0.865639 -1.387538  0.506303      -0.911583         0.326305          1               0

[6700 rows x 11 columns]
[[1]
 [0]
 [0]
 ...
 [0]
 [0]
 [1]]
Modelimizi oluşturmaya başlayabiliriz. Bunun için keras.model.Sequential() sınıfından bir obje oluşturuyoruz. Bu sınıfın add methodunu kullanarak katmana eklemeler yapacağız. Burada 11 bağımsız sütunumuz ve 1 bağımlı sütunumuz var. Bu nedenle ilk katman 11 nöründan, çıkış katmanı ise 1 nörondan oluşacak. Arada iki gizli katman kullanacağız bu katmanlar da 6 nörondan oluşacak. Çıkış katmanında sigmoid aktivasyonu, diğer katmanlarda relu aktivasyonunu kullanacağız:


model = Sequential()
model.add(Dense(6, activation='relu', input_dim=11))
model.add(Dense(6, activation='relu'))
model.add(Dense(1, activation='sigmoid'))
Modeli tamamlamak için compile methodunu kullanıyoruz. Parametre olarak bir optimizer, loss fonksiyonu ve metrikler vermeliyiz. Burada adam optimizer, binary_crossentropy loss fonksiyonu ve accuracy metriğini kullanacağız:


model.compile(optimizer='adam', loss='binary_crossentropy', metrics=['accuracy'])
Son olarak modeli fit edelim. Ve fit ettiğimiz modeli kullanarak x_test için tahminler yapalım. Confusion matrix ile sonuçları görelim. Modelin fit methodunu kullanırken epoch değerini 50 vereceğiz:


model.fit(x_train, y_train, epochs=50)
res = model.predict(x_test)
res = (res > .5)
cm = confusion_matrix(res, y_test)
print(cm)

>>

  32/6700 [..............................] - ETA: 3:28 - loss: 0.7133 - acc: 0.4062
1184/6700 [====>.........................] - ETA: 4s - loss: 0.6492 - acc: 0.5929  
2528/6700 [==========>...................] - ETA: 1s - loss: 0.6115 - acc: 0.6990
3936/6700 [================>.............] - ETA: 0s - loss: 0.5937 - acc: 0.7251
5120/6700 [=====================>........] - ETA: 0s - loss: 0.5737 - acc: 0.7434
6464/6700 [===========================>..] - ETA: 0s - loss: 0.5546 - acc: 0.7562
6700/6700 [==============================] - 1s 196us/step - loss: 0.5523 - acc: 0.7584
Epoch 2/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4160 - acc: 0.8438
1184/6700 [====>.........................] - ETA: 0s - loss: 0.4582 - acc: 0.8083
2560/6700 [==========>...................] - ETA: 0s - loss: 0.4679 - acc: 0.8008
3776/6700 [===============>..............] - ETA: 0s - loss: 0.4725 - acc: 0.7971
5152/6700 [======================>.......] - ETA: 0s - loss: 0.4751 - acc: 0.7937
6496/6700 [============================>.] - ETA: 0s - loss: 0.4656 - acc: 0.7983
6700/6700 [==============================] - 0s 47us/step - loss: 0.4650 - acc: 0.7979
Epoch 3/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4907 - acc: 0.7188
1184/6700 [====>.........................] - ETA: 0s - loss: 0.4647 - acc: 0.7829
2624/6700 [==========>...................] - ETA: 0s - loss: 0.4641 - acc: 0.7877
3744/6700 [===============>..............] - ETA: 0s - loss: 0.4557 - acc: 0.7914
5088/6700 [=====================>........] - ETA: 0s - loss: 0.4487 - acc: 0.7954
6368/6700 [===========================>..] - ETA: 0s - loss: 0.4447 - acc: 0.7976
6700/6700 [==============================] - 0s 49us/step - loss: 0.4430 - acc: 0.7982
Epoch 4/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4417 - acc: 0.7812
1088/6700 [===>..........................] - ETA: 0s - loss: 0.4233 - acc: 0.8051
2400/6700 [=========>....................] - ETA: 0s - loss: 0.4352 - acc: 0.8021
3616/6700 [===============>..............] - ETA: 0s - loss: 0.4343 - acc: 0.7987
4864/6700 [====================>.........] - ETA: 0s - loss: 0.4314 - acc: 0.8014
5952/6700 [=========================>....] - ETA: 0s - loss: 0.4351 - acc: 0.7987
6700/6700 [==============================] - 0s 51us/step - loss: 0.4331 - acc: 0.8000
Epoch 5/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4833 - acc: 0.7500
 928/6700 [===>..........................] - ETA: 0s - loss: 0.4510 - acc: 0.7866
2144/6700 [========>.....................] - ETA: 0s - loss: 0.4345 - acc: 0.7966
3424/6700 [==============>...............] - ETA: 0s - loss: 0.4324 - acc: 0.7982
4736/6700 [====================>.........] - ETA: 0s - loss: 0.4262 - acc: 0.8022
6112/6700 [==========================>...] - ETA: 0s - loss: 0.4271 - acc: 0.8032
6700/6700 [==============================] - 0s 49us/step - loss: 0.4274 - acc: 0.8034
Epoch 6/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4193 - acc: 0.8125
 992/6700 [===>..........................] - ETA: 0s - loss: 0.4339 - acc: 0.7964
2400/6700 [=========>....................] - ETA: 0s - loss: 0.4298 - acc: 0.7987
3776/6700 [===============>..............] - ETA: 0s - loss: 0.4227 - acc: 0.8046
5088/6700 [=====================>........] - ETA: 0s - loss: 0.4203 - acc: 0.8027
6176/6700 [==========================>...] - ETA: 0s - loss: 0.4239 - acc: 0.8002
6700/6700 [==============================] - 0s 49us/step - loss: 0.4228 - acc: 0.8015
Epoch 7/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4178 - acc: 0.7500
1088/6700 [===>..........................] - ETA: 0s - loss: 0.4118 - acc: 0.7987
2432/6700 [=========>....................] - ETA: 0s - loss: 0.4128 - acc: 0.8022
3712/6700 [===============>..............] - ETA: 0s - loss: 0.4201 - acc: 0.8015
4960/6700 [=====================>........] - ETA: 0s - loss: 0.4170 - acc: 0.8036
6368/6700 [===========================>..] - ETA: 0s - loss: 0.4187 - acc: 0.8028
6700/6700 [==============================] - 0s 49us/step - loss: 0.4175 - acc: 0.8028
Epoch 8/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4138 - acc: 0.8750
1504/6700 [=====>........................] - ETA: 0s - loss: 0.4134 - acc: 0.8078
2688/6700 [===========>..................] - ETA: 0s - loss: 0.4098 - acc: 0.8058
4000/6700 [================>.............] - ETA: 0s - loss: 0.4024 - acc: 0.8097
5376/6700 [=======================>......] - ETA: 0s - loss: 0.4096 - acc: 0.8043
6700/6700 [==============================] - 0s 44us/step - loss: 0.4117 - acc: 0.8061
Epoch 9/50

  32/6700 [..............................] - ETA: 0s - loss: 0.5383 - acc: 0.7188
1184/6700 [====>.........................] - ETA: 0s - loss: 0.4057 - acc: 0.8193
2560/6700 [==========>...................] - ETA: 0s - loss: 0.3997 - acc: 0.8148
4000/6700 [================>.............] - ETA: 0s - loss: 0.4082 - acc: 0.8165
5536/6700 [=======================>......] - ETA: 0s - loss: 0.4099 - acc: 0.8148
6700/6700 [==============================] - 0s 44us/step - loss: 0.4043 - acc: 0.8203
Epoch 10/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2466 - acc: 0.9375
1184/6700 [====>.........................] - ETA: 0s - loss: 0.3922 - acc: 0.8218
2656/6700 [==========>...................] - ETA: 0s - loss: 0.4010 - acc: 0.8159
4192/6700 [=================>............] - ETA: 0s - loss: 0.4052 - acc: 0.8201
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3998 - acc: 0.8242
6700/6700 [==============================] - 0s 44us/step - loss: 0.3957 - acc: 0.8282
Epoch 11/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2741 - acc: 0.9062
1312/6700 [====>.........................] - ETA: 0s - loss: 0.3646 - acc: 0.8384
2816/6700 [===========>..................] - ETA: 0s - loss: 0.3865 - acc: 0.8356
4288/6700 [==================>...........] - ETA: 0s - loss: 0.3912 - acc: 0.8337
5664/6700 [========================>.....] - ETA: 0s - loss: 0.3893 - acc: 0.8328
6700/6700 [==============================] - 0s 44us/step - loss: 0.3866 - acc: 0.8343
Epoch 12/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3872 - acc: 0.8125
1504/6700 [=====>........................] - ETA: 0s - loss: 0.3673 - acc: 0.8378
2976/6700 [============>.................] - ETA: 0s - loss: 0.3595 - acc: 0.8431
4384/6700 [==================>...........] - ETA: 0s - loss: 0.3753 - acc: 0.8374
5792/6700 [========================>.....] - ETA: 0s - loss: 0.3768 - acc: 0.8398
6700/6700 [==============================] - 0s 42us/step - loss: 0.3788 - acc: 0.8394
Epoch 13/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2804 - acc: 0.9062
1376/6700 [=====>........................] - ETA: 0s - loss: 0.3837 - acc: 0.8372
2752/6700 [===========>..................] - ETA: 0s - loss: 0.3649 - acc: 0.8470
4032/6700 [=================>............] - ETA: 0s - loss: 0.3592 - acc: 0.8504
5536/6700 [=======================>......] - ETA: 0s - loss: 0.3722 - acc: 0.8441
6700/6700 [==============================] - 0s 44us/step - loss: 0.3720 - acc: 0.8454
Epoch 14/50

  32/6700 [..............................] - ETA: 0s - loss: 0.1979 - acc: 0.9688
1312/6700 [====>.........................] - ETA: 0s - loss: 0.3619 - acc: 0.8491
2816/6700 [===========>..................] - ETA: 0s - loss: 0.3638 - acc: 0.8491
4384/6700 [==================>...........] - ETA: 0s - loss: 0.3675 - acc: 0.8481
5824/6700 [=========================>....] - ETA: 0s - loss: 0.3625 - acc: 0.8511
6700/6700 [==============================] - 0s 42us/step - loss: 0.3672 - acc: 0.8497
Epoch 15/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3047 - acc: 0.9062
1152/6700 [====>.........................] - ETA: 0s - loss: 0.3416 - acc: 0.8646
2624/6700 [==========>...................] - ETA: 0s - loss: 0.3437 - acc: 0.8613
4192/6700 [=================>............] - ETA: 0s - loss: 0.3450 - acc: 0.8585
5536/6700 [=======================>......] - ETA: 0s - loss: 0.3558 - acc: 0.8557
6700/6700 [==============================] - 0s 44us/step - loss: 0.3633 - acc: 0.8533
Epoch 16/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3139 - acc: 0.9375
1312/6700 [====>.........................] - ETA: 0s - loss: 0.3569 - acc: 0.8720
2816/6700 [===========>..................] - ETA: 0s - loss: 0.3584 - acc: 0.8636
4160/6700 [=================>............] - ETA: 0s - loss: 0.3589 - acc: 0.8587
5600/6700 [========================>.....] - ETA: 0s - loss: 0.3591 - acc: 0.8559
6700/6700 [==============================] - 0s 44us/step - loss: 0.3608 - acc: 0.8564
Epoch 17/50

  32/6700 [..............................] - ETA: 0s - loss: 0.1744 - acc: 1.0000
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3574 - acc: 0.8577
2592/6700 [==========>...................] - ETA: 0s - loss: 0.3561 - acc: 0.8654
4032/6700 [=================>............] - ETA: 0s - loss: 0.3629 - acc: 0.8601
5568/6700 [=======================>......] - ETA: 0s - loss: 0.3655 - acc: 0.8540
6700/6700 [==============================] - 0s 44us/step - loss: 0.3589 - acc: 0.8560
Epoch 18/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2971 - acc: 0.8438
1344/6700 [=====>........................] - ETA: 0s - loss: 0.3383 - acc: 0.8609
2848/6700 [===========>..................] - ETA: 0s - loss: 0.3502 - acc: 0.8560
4416/6700 [==================>...........] - ETA: 0s - loss: 0.3527 - acc: 0.8585
5824/6700 [=========================>....] - ETA: 0s - loss: 0.3548 - acc: 0.8556
6700/6700 [==============================] - 0s 42us/step - loss: 0.3575 - acc: 0.8545
Epoch 19/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4584 - acc: 0.7812
1056/6700 [===>..........................] - ETA: 0s - loss: 0.3886 - acc: 0.8475
2496/6700 [==========>...................] - ETA: 0s - loss: 0.3706 - acc: 0.8514
4064/6700 [=================>............] - ETA: 0s - loss: 0.3623 - acc: 0.8561
5408/6700 [=======================>......] - ETA: 0s - loss: 0.3590 - acc: 0.8569
6700/6700 [==============================] - 0s 44us/step - loss: 0.3561 - acc: 0.8564
Epoch 20/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3773 - acc: 0.7812
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3895 - acc: 0.8396
2496/6700 [==========>...................] - ETA: 0s - loss: 0.3676 - acc: 0.8546
3872/6700 [================>.............] - ETA: 0s - loss: 0.3598 - acc: 0.8559
5248/6700 [======================>.......] - ETA: 0s - loss: 0.3571 - acc: 0.8533
6700/6700 [==============================] - 0s 44us/step - loss: 0.3551 - acc: 0.8566
Epoch 21/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3073 - acc: 0.8125
1120/6700 [====>.........................] - ETA: 0s - loss: 0.3541 - acc: 0.8679
2496/6700 [==========>...................] - ETA: 0s - loss: 0.3496 - acc: 0.8638
3936/6700 [================>.............] - ETA: 0s - loss: 0.3574 - acc: 0.8582
5472/6700 [=======================>......] - ETA: 0s - loss: 0.3563 - acc: 0.8580
6700/6700 [==============================] - 0s 44us/step - loss: 0.3539 - acc: 0.8578
Epoch 22/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4210 - acc: 0.8750
1280/6700 [====>.........................] - ETA: 0s - loss: 0.3475 - acc: 0.8500
2752/6700 [===========>..................] - ETA: 0s - loss: 0.3494 - acc: 0.8561
4288/6700 [==================>...........] - ETA: 0s - loss: 0.3434 - acc: 0.8622
5728/6700 [========================>.....] - ETA: 0s - loss: 0.3448 - acc: 0.8628
6700/6700 [==============================] - 0s 44us/step - loss: 0.3533 - acc: 0.8587
Epoch 23/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3537 - acc: 0.9062
1376/6700 [=====>........................] - ETA: 0s - loss: 0.3248 - acc: 0.8794
2944/6700 [============>.................] - ETA: 0s - loss: 0.3510 - acc: 0.8607
4256/6700 [==================>...........] - ETA: 0s - loss: 0.3482 - acc: 0.8611
5184/6700 [======================>.......] - ETA: 0s - loss: 0.3516 - acc: 0.8573
6240/6700 [==========================>...] - ETA: 0s - loss: 0.3517 - acc: 0.8583
6700/6700 [==============================] - 0s 46us/step - loss: 0.3528 - acc: 0.8579
Epoch 24/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3944 - acc: 0.8750
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3392 - acc: 0.8668
2304/6700 [=========>....................] - ETA: 0s - loss: 0.3505 - acc: 0.8581
3296/6700 [=============>................] - ETA: 0s - loss: 0.3451 - acc: 0.8613
4672/6700 [===================>..........] - ETA: 0s - loss: 0.3474 - acc: 0.8617
6208/6700 [==========================>...] - ETA: 0s - loss: 0.3515 - acc: 0.8574
6700/6700 [==============================] - 0s 44us/step - loss: 0.3518 - acc: 0.8573
Epoch 25/50

  32/6700 [..............................] - ETA: 0s - loss: 0.6017 - acc: 0.7812
1312/6700 [====>.........................] - ETA: 0s - loss: 0.3445 - acc: 0.8605
2752/6700 [===========>..................] - ETA: 0s - loss: 0.3437 - acc: 0.8641
4288/6700 [==================>...........] - ETA: 0s - loss: 0.3460 - acc: 0.8624
5728/6700 [========================>.....] - ETA: 0s - loss: 0.3445 - acc: 0.8621
6700/6700 [==============================] - 0s 44us/step - loss: 0.3518 - acc: 0.8587
Epoch 26/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4612 - acc: 0.7812
1376/6700 [=====>........................] - ETA: 0s - loss: 0.3452 - acc: 0.8561
2912/6700 [============>.................] - ETA: 0s - loss: 0.3553 - acc: 0.8565
4192/6700 [=================>............] - ETA: 0s - loss: 0.3500 - acc: 0.8585
4928/6700 [=====================>........] - ETA: 0s - loss: 0.3500 - acc: 0.8588
5792/6700 [========================>.....] - ETA: 0s - loss: 0.3526 - acc: 0.8584
6688/6700 [============================>.] - ETA: 0s - loss: 0.3510 - acc: 0.8599
6700/6700 [==============================] - 0s 56us/step - loss: 0.3510 - acc: 0.8600
Epoch 27/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3407 - acc: 0.8750
 960/6700 [===>..........................] - ETA: 0s - loss: 0.3350 - acc: 0.8635
1792/6700 [=======>......................] - ETA: 0s - loss: 0.3427 - acc: 0.8683
2720/6700 [===========>..................] - ETA: 0s - loss: 0.3528 - acc: 0.8599
3712/6700 [===============>..............] - ETA: 0s - loss: 0.3545 - acc: 0.8578
4704/6700 [====================>.........] - ETA: 0s - loss: 0.3542 - acc: 0.8580
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3528 - acc: 0.8588
6560/6700 [============================>.] - ETA: 0s - loss: 0.3503 - acc: 0.8596
6700/6700 [==============================] - 0s 65us/step - loss: 0.3503 - acc: 0.8593
Epoch 28/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2623 - acc: 0.9062
 736/6700 [==>...........................] - ETA: 0s - loss: 0.3815 - acc: 0.8410
1664/6700 [======>.......................] - ETA: 0s - loss: 0.3717 - acc: 0.8462
2400/6700 [=========>....................] - ETA: 0s - loss: 0.3724 - acc: 0.8496
3232/6700 [=============>................] - ETA: 0s - loss: 0.3661 - acc: 0.8552
4448/6700 [==================>...........] - ETA: 0s - loss: 0.3581 - acc: 0.8570
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3510 - acc: 0.8596
6700/6700 [==============================] - 0s 59us/step - loss: 0.3499 - acc: 0.8601
Epoch 29/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4307 - acc: 0.8125
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3473 - acc: 0.8651
2688/6700 [===========>..................] - ETA: 0s - loss: 0.3501 - acc: 0.8605
4160/6700 [=================>............] - ETA: 0s - loss: 0.3518 - acc: 0.8579
5504/6700 [=======================>......] - ETA: 0s - loss: 0.3508 - acc: 0.8592
6700/6700 [==============================] - 0s 43us/step - loss: 0.3495 - acc: 0.8594
Epoch 30/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3591 - acc: 0.8438
1344/6700 [=====>........................] - ETA: 0s - loss: 0.3429 - acc: 0.8624
2784/6700 [===========>..................] - ETA: 0s - loss: 0.3434 - acc: 0.8675
4192/6700 [=================>............] - ETA: 0s - loss: 0.3454 - acc: 0.8609
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3476 - acc: 0.8585
6700/6700 [==============================] - 0s 42us/step - loss: 0.3488 - acc: 0.8588
Epoch 31/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2483 - acc: 0.8750
1184/6700 [====>.........................] - ETA: 0s - loss: 0.3218 - acc: 0.8742
2560/6700 [==========>...................] - ETA: 0s - loss: 0.3340 - acc: 0.8672
3968/6700 [================>.............] - ETA: 0s - loss: 0.3387 - acc: 0.8649
5440/6700 [=======================>......] - ETA: 0s - loss: 0.3465 - acc: 0.8603
6700/6700 [==============================] - 0s 44us/step - loss: 0.3484 - acc: 0.8596
Epoch 32/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2907 - acc: 0.8750
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3464 - acc: 0.8536
2720/6700 [===========>..................] - ETA: 0s - loss: 0.3520 - acc: 0.8570
4256/6700 [==================>...........] - ETA: 0s - loss: 0.3508 - acc: 0.8616
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3460 - acc: 0.8633
6700/6700 [==============================] - 0s 44us/step - loss: 0.3484 - acc: 0.8601
Epoch 33/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3235 - acc: 0.9062
1344/6700 [=====>........................] - ETA: 0s - loss: 0.3726 - acc: 0.8423
2848/6700 [===========>..................] - ETA: 0s - loss: 0.3543 - acc: 0.8592
4288/6700 [==================>...........] - ETA: 0s - loss: 0.3480 - acc: 0.8619
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3430 - acc: 0.8624
6700/6700 [==============================] - 0s 44us/step - loss: 0.3482 - acc: 0.8603
Epoch 34/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2249 - acc: 0.9688
1472/6700 [=====>........................] - ETA: 0s - loss: 0.3437 - acc: 0.8648
2944/6700 [============>.................] - ETA: 0s - loss: 0.3467 - acc: 0.8624
4320/6700 [==================>...........] - ETA: 0s - loss: 0.3480 - acc: 0.8618
5760/6700 [========================>.....] - ETA: 0s - loss: 0.3525 - acc: 0.8589
6700/6700 [==============================] - 0s 42us/step - loss: 0.3476 - acc: 0.8612
Epoch 35/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3254 - acc: 0.9062
1312/6700 [====>.........................] - ETA: 0s - loss: 0.3556 - acc: 0.8575
2656/6700 [==========>...................] - ETA: 0s - loss: 0.3624 - acc: 0.8505
4064/6700 [=================>............] - ETA: 0s - loss: 0.3533 - acc: 0.8570
5568/6700 [=======================>......] - ETA: 0s - loss: 0.3484 - acc: 0.8614
6700/6700 [==============================] - 0s 44us/step - loss: 0.3478 - acc: 0.8609
Epoch 36/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3011 - acc: 0.9062
1056/6700 [===>..........................] - ETA: 0s - loss: 0.3678 - acc: 0.8542
2176/6700 [========>.....................] - ETA: 0s - loss: 0.3510 - acc: 0.8598
3424/6700 [==============>...............] - ETA: 0s - loss: 0.3480 - acc: 0.8592
4608/6700 [===================>..........] - ETA: 0s - loss: 0.3492 - acc: 0.8583
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3510 - acc: 0.8574
6624/6700 [============================>.] - ETA: 0s - loss: 0.3472 - acc: 0.8604
6700/6700 [==============================] - 0s 47us/step - loss: 0.3471 - acc: 0.8604
Epoch 37/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3465 - acc: 0.9062
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3537 - acc: 0.8734
2432/6700 [=========>....................] - ETA: 0s - loss: 0.3683 - acc: 0.8549
3552/6700 [==============>...............] - ETA: 0s - loss: 0.3557 - acc: 0.8595
4576/6700 [===================>..........] - ETA: 0s - loss: 0.3471 - acc: 0.8634
5600/6700 [========================>.....] - ETA: 0s - loss: 0.3451 - acc: 0.8634
6700/6700 [==============================] - 0s 46us/step - loss: 0.3468 - acc: 0.8613
Epoch 38/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2656 - acc: 0.9062
1280/6700 [====>.........................] - ETA: 0s - loss: 0.3123 - acc: 0.8766
2272/6700 [=========>....................] - ETA: 0s - loss: 0.3340 - acc: 0.8715
3264/6700 [=============>................] - ETA: 0s - loss: 0.3468 - acc: 0.8637
4416/6700 [==================>...........] - ETA: 0s - loss: 0.3474 - acc: 0.8644
5664/6700 [========================>.....] - ETA: 0s - loss: 0.3463 - acc: 0.8644
6700/6700 [==============================] - 0s 45us/step - loss: 0.3468 - acc: 0.8634
Epoch 39/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2119 - acc: 0.9062
1056/6700 [===>..........................] - ETA: 0s - loss: 0.3400 - acc: 0.8608
2048/6700 [========>.....................] - ETA: 0s - loss: 0.3510 - acc: 0.8550
3296/6700 [=============>................] - ETA: 0s - loss: 0.3604 - acc: 0.8547
4864/6700 [====================>.........] - ETA: 0s - loss: 0.3477 - acc: 0.8612
6336/6700 [===========================>..] - ETA: 0s - loss: 0.3441 - acc: 0.8640
6700/6700 [==============================] - 0s 45us/step - loss: 0.3461 - acc: 0.8625
Epoch 40/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2898 - acc: 0.8438
1376/6700 [=====>........................] - ETA: 0s - loss: 0.3579 - acc: 0.8452
2912/6700 [============>.................] - ETA: 0s - loss: 0.3500 - acc: 0.8561
4448/6700 [==================>...........] - ETA: 0s - loss: 0.3477 - acc: 0.8597
5856/6700 [=========================>....] - ETA: 0s - loss: 0.3472 - acc: 0.8610
6700/6700 [==============================] - 0s 42us/step - loss: 0.3461 - acc: 0.8622
Epoch 41/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4696 - acc: 0.7500
1152/6700 [====>.........................] - ETA: 0s - loss: 0.3453 - acc: 0.8628
2592/6700 [==========>...................] - ETA: 0s - loss: 0.3585 - acc: 0.8561
4160/6700 [=================>............] - ETA: 0s - loss: 0.3498 - acc: 0.8601
5536/6700 [=======================>......] - ETA: 0s - loss: 0.3470 - acc: 0.8613
6700/6700 [==============================] - 0s 44us/step - loss: 0.3458 - acc: 0.8622
Epoch 42/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3067 - acc: 0.8125
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3522 - acc: 0.8627
2624/6700 [==========>...................] - ETA: 0s - loss: 0.3436 - acc: 0.8636
3968/6700 [================>.............] - ETA: 0s - loss: 0.3424 - acc: 0.8629
5376/6700 [=======================>......] - ETA: 0s - loss: 0.3413 - acc: 0.8657
6656/6700 [============================>.] - ETA: 0s - loss: 0.3459 - acc: 0.8633
6700/6700 [==============================] - 0s 47us/step - loss: 0.3459 - acc: 0.8627
Epoch 43/50

  32/6700 [..............................] - ETA: 0s - loss: 0.2163 - acc: 0.9375
1344/6700 [=====>........................] - ETA: 0s - loss: 0.3449 - acc: 0.8698
2624/6700 [==========>...................] - ETA: 0s - loss: 0.3430 - acc: 0.8639
4032/6700 [=================>............] - ETA: 0s - loss: 0.3460 - acc: 0.8621
5536/6700 [=======================>......] - ETA: 0s - loss: 0.3488 - acc: 0.8620
6700/6700 [==============================] - 0s 44us/step - loss: 0.3453 - acc: 0.8633
Epoch 44/50

  32/6700 [..............................] - ETA: 0s - loss: 0.6262 - acc: 0.7500
1248/6700 [====>.........................] - ETA: 0s - loss: 0.3524 - acc: 0.8614
2688/6700 [===========>..................] - ETA: 0s - loss: 0.3646 - acc: 0.8534
4224/6700 [=================>............] - ETA: 0s - loss: 0.3484 - acc: 0.8601
5344/6700 [======================>.......] - ETA: 0s - loss: 0.3474 - acc: 0.8604
6272/6700 [===========================>..] - ETA: 0s - loss: 0.3472 - acc: 0.8616
6700/6700 [==============================] - 0s 49us/step - loss: 0.3452 - acc: 0.8621
Epoch 45/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3717 - acc: 0.8125
1248/6700 [====>.........................] - ETA: 0s - loss: 0.3647 - acc: 0.8662
2752/6700 [===========>..................] - ETA: 0s - loss: 0.3482 - acc: 0.8641
4128/6700 [=================>............] - ETA: 0s - loss: 0.3407 - acc: 0.8670
5472/6700 [=======================>......] - ETA: 0s - loss: 0.3447 - acc: 0.8635
6700/6700 [==============================] - 0s 44us/step - loss: 0.3448 - acc: 0.8639
Epoch 46/50

  32/6700 [..............................] - ETA: 2s - loss: 0.3305 - acc: 0.8750
1056/6700 [===>..........................] - ETA: 0s - loss: 0.3391 - acc: 0.8693
2208/6700 [========>.....................] - ETA: 0s - loss: 0.3355 - acc: 0.8696
3232/6700 [=============>................] - ETA: 0s - loss: 0.3331 - acc: 0.8713
4256/6700 [==================>...........] - ETA: 0s - loss: 0.3403 - acc: 0.8665
5472/6700 [=======================>......] - ETA: 0s - loss: 0.3427 - acc: 0.8633
6656/6700 [============================>.] - ETA: 0s - loss: 0.3446 - acc: 0.8636
6700/6700 [==============================] - 0s 49us/step - loss: 0.3445 - acc: 0.8639
Epoch 47/50

  32/6700 [..............................] - ETA: 0s - loss: 0.1475 - acc: 1.0000
1216/6700 [====>.........................] - ETA: 0s - loss: 0.3338 - acc: 0.8742
2688/6700 [===========>..................] - ETA: 0s - loss: 0.3317 - acc: 0.8709
4224/6700 [=================>............] - ETA: 0s - loss: 0.3395 - acc: 0.8684
5696/6700 [========================>.....] - ETA: 0s - loss: 0.3448 - acc: 0.8648
6700/6700 [==============================] - 0s 43us/step - loss: 0.3443 - acc: 0.8637
Epoch 48/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4041 - acc: 0.8438
1280/6700 [====>.........................] - ETA: 0s - loss: 0.4035 - acc: 0.8258
2816/6700 [===========>..................] - ETA: 0s - loss: 0.3636 - acc: 0.8501
4256/6700 [==================>...........] - ETA: 0s - loss: 0.3520 - acc: 0.8569
5632/6700 [========================>.....] - ETA: 0s - loss: 0.3484 - acc: 0.8604
6700/6700 [==============================] - 0s 44us/step - loss: 0.3441 - acc: 0.8631
Epoch 49/50

  32/6700 [..............................] - ETA: 0s - loss: 0.3398 - acc: 0.8438
1376/6700 [=====>........................] - ETA: 0s - loss: 0.3407 - acc: 0.8699
2464/6700 [==========>...................] - ETA: 0s - loss: 0.3520 - acc: 0.8588
3520/6700 [==============>...............] - ETA: 0s - loss: 0.3504 - acc: 0.8622
4576/6700 [===================>..........] - ETA: 0s - loss: 0.3510 - acc: 0.8621
5760/6700 [========================>.....] - ETA: 0s - loss: 0.3479 - acc: 0.8634
6700/6700 [==============================] - 0s 46us/step - loss: 0.3441 - acc: 0.8634
Epoch 50/50

  32/6700 [..............................] - ETA: 0s - loss: 0.4607 - acc: 0.9062
 992/6700 [===>..........................] - ETA: 0s - loss: 0.3248 - acc: 0.8679
1760/6700 [======>.......................] - ETA: 0s - loss: 0.3334 - acc: 0.8653
2336/6700 [=========>....................] - ETA: 0s - loss: 0.3277 - acc: 0.8694
3072/6700 [============>.................] - ETA: 0s - loss: 0.3356 - acc: 0.8639
3744/6700 [===============>..............] - ETA: 0s - loss: 0.3377 - acc: 0.8638
4512/6700 [===================>..........] - ETA: 0s - loss: 0.3461 - acc: 0.8606
5440/6700 [=======================>......] - ETA: 0s - loss: 0.3423 - acc: 0.8629
6624/6700 [============================>.] - ETA: 0s - loss: 0.3440 - acc: 0.8629
6700/6700 [==============================] - 0s 70us/step - loss: 0.3438 - acc: 0.8630

[[2480  341]
 [ 137  342]]


Sonraki Yazı: Principal Component Analysis
Yorumlar

Henüz bir yorum bulunmuyor.
Yorum bırakın