Problema de clasificación

# Importaciones necesarias
import numpy as np
import seaborn as sns
import pandas as pd
import matplotlib.pyplot as plt

# Cargamos los datos
df = pd.read_csv('titanic.csv')
df.head(2)

	survived	pclass	age	sibsp	parch	fare	male	age_was_missing	embarked_from_cherbourg	embarked_from_queenstown	embarked_from_southampton
0	0	3	22.0	1	0	7.2500	1	False	0	0	1
1	1	1	38.0	1	0	71.2833	0	False	1	0	0

df['age_was_missing'].unique()

array([False,  True])

# numero de filas por numero de columnas
df.shape

(891, 11)

La variable objetivo corresponde a la columna survived que tiene marcado el 1 si el pasajero sobrevivió y 0 caso contrario. El resto de columnas coresponden a las características.

De tal manera, nuestro objetivo será predecir si el pasajero sobrevivió o no.

Notemos antes que de los pasajeros, sobrevivieron más mujeres (0) que hombres (1)

plt.figure(figsize=(4,4))
sns.scatterplot(data=df, x="age", y="male", hue='survived')
plt.title('Sexo vs Edad: Sobrevivientes')
plt.show()

Continuando, nuestra red tendrá 10 nodos de entrada y un nodo de salida; la capa de salida tendrá asociada la función sigmoide, la cual nos arrojará 0 para números obtenido menores al umbral (por defecto el umbral vale 0.5), y 1 para números mayores o iguales al umbral. Crearemos dos capas ocultas y le asociaremos la función de activación relu

Convertiremos nuestro dataframe a un array

df.columns

Index(['survived', 'pclass', 'age', 'sibsp', 'parch', 'fare', 'male',
       'age_was_missing', 'embarked_from_cherbourg',
       'embarked_from_queenstown', 'embarked_from_southampton'],
      dtype='object')

# Convertimos el dataframe a un array
df_np = np.array(df)

# Veamos que la columna 7 tiene entradas booleanas
df_np[0:3,7].reshape(3,1)

array([[False],
       [False],
       [False]], dtype=object)

# Cambiaremos el valor de True a 1 y el de False a 0
filtrado = np.where(df_np[:,7] , 1, 0)

# A la columna 7 le asignamos los valores obtenidos
# anteriormente
df_np[:,7] = filtrado

Procedemos a dividir los datos:

df_np[:, 0]

array([0, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 1, 0, 1, 0, 1, 0, 1,
       1, 1, 0, 1, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1,
       1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 1, 0, 0, 1, 0, 0, 0, 1,
       1, 0, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 1, 1, 0, 1, 1, 0, 0,
       1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1,
       0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 1, 1, 0, 0, 0,
       0, 1, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0,
       0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 1, 1, 1, 1, 0, 0,
       1, 0, 0, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 1, 0, 0, 0, 1, 1, 0, 1, 0,
       1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0, 1, 0, 0, 0, 1, 0, 0, 0, 1,
       0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 1, 0, 0,
       0, 0, 0, 1, 1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 1, 0, 0,
       1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1,
       0, 1, 1, 1, 0, 0, 0, 1, 1, 0, 1, 1, 0, 0, 1, 1, 0, 1, 0, 1, 1, 1,
       1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0,
       0, 0, 0, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0,
       0, 1, 1, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 1, 0, 1, 1, 1, 0, 1, 1, 0,
       0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 1, 0, 1, 1,
       0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, 0,
       1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0, 1, 0, 1, 1, 0, 1, 0,
       0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1,
       1, 0, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 1, 0,
       1, 1, 0, 1, 1, 0, 1, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1, 0, 0, 1, 0,
       0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 1, 1, 0, 0, 1, 0, 0, 1, 1, 0, 1,
       1, 0, 0, 1, 1, 0, 1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,
       1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 0, 0, 0, 1, 0, 1, 0, 0, 0, 1, 0, 0,
       0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 1, 0, 0, 1, 0, 0, 1,
       0, 0, 1, 0, 0, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 0, 1, 0, 0,
       0, 0, 0, 1, 0, 1, 1, 1, 0, 1, 0, 1, 0, 1, 0, 1, 0, 0, 0, 0, 0, 0,
       1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 1, 0, 1, 0, 1,
       0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0,
       0, 0, 1, 1, 1, 1, 1, 0, 1, 0, 0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 1, 0,
       1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 0, 1, 0, 0, 1,
       0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 1, 0, 0, 0, 0,
       0, 0, 0, 0, 1, 0, 0, 1, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0,
       0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0,
       0, 0, 0, 0, 0, 0, 1, 1, 0, 1, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 1,
       0, 0, 1, 1, 0, 0, 1, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 1, 0, 1, 1, 1,
       1, 0, 0, 0, 1, 0, 0, 1, 1, 0, 0, 1, 0, 1, 0, 0, 1, 1, 0, 0, 0, 1,
       1, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0], dtype=object)

from sklearn.model_selection import train_test_split

# Las columnas de la 1 a la ultima son nuestras
# caracteristicas
X = df_np[:, 1:11]
X = np.asarray(X).astype('float32')

# La variable objetivo es la primera columna
y = df_np[:, 0].reshape(891,1)
y = np.asarray(y).astype('float32')

# División de los datos
X_train, X_test, y_train, y_test = train_test_split(
                                        X,
                                        y,
                                        train_size   = 0.75,
                                        random_state = 123)

print(X_train.shape)
print(X_test.shape)
print(y_train.shape)
print(y_test.shape)

(668, 10)
(223, 10)
(668, 1)
(223, 1)

Ahora creamos nuestro modelo en keras:

# Realizamos las importaciones necesarias
from keras.models import Sequential
from keras.layers import Dense

# Fijamos una semilla
np.random.seed(123)

# Instanciamos el modelo
model = Sequential()

# Capa de entrada de 10 neuronas y capa oculta
# de 70 neuronas
model.add(Dense(50, input_shape=(10,) , activation='relu'))



# Capa de salida
model.add(Dense(1, activation='sigmoid'))

# Compilacion del modelo.
# El optimizador: adam 
# Funcion de perdida: binary_crossentropy
model.compile(optimizer='adam', 
              loss='binary_crossentropy', 
              metrics=['accuracy'])

# Vemos un resumen sobre nuestro modelo
model.summary()

Model: "sequential_20"
_________________________________________________________________
 Layer (type)                Output Shape              Param #   
=================================================================
 dense_55 (Dense)            (None, 50)                550       
                                                                 
 dense_56 (Dense)            (None, 1)                 51        
                                                                 
=================================================================
Total params: 601
Trainable params: 601
Non-trainable params: 0
_________________________________________________________________

# Ajustamos el modelo con los datos de entrenamiento
history = model.fit(X_train, y_train,
          epochs=100,
          validation_data=(X_test, y_test))

Epoch 1/100
21/21 [==============================] - 1s 8ms/step - loss: 3.3755 - accuracy: 0.3817 - val_loss: 1.1949 - val_accuracy: 0.3139
Epoch 2/100
21/21 [==============================] - 0s 3ms/step - loss: 0.8643 - accuracy: 0.6108 - val_loss: 0.6134 - val_accuracy: 0.7130
Epoch 3/100
21/21 [==============================] - 0s 3ms/step - loss: 0.6225 - accuracy: 0.7081 - val_loss: 0.6028 - val_accuracy: 0.7265
Epoch 4/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5840 - accuracy: 0.7081 - val_loss: 0.5845 - val_accuracy: 0.7220
Epoch 5/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5734 - accuracy: 0.7141 - val_loss: 0.5841 - val_accuracy: 0.7265
Epoch 6/100
21/21 [==============================] - 0s 2ms/step - loss: 0.5678 - accuracy: 0.7275 - val_loss: 0.5683 - val_accuracy: 0.7534
Epoch 7/100
21/21 [==============================] - 0s 2ms/step - loss: 0.5588 - accuracy: 0.7275 - val_loss: 0.5544 - val_accuracy: 0.7534
Epoch 8/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5516 - accuracy: 0.7395 - val_loss: 0.5614 - val_accuracy: 0.7309
Epoch 9/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5485 - accuracy: 0.7395 - val_loss: 0.5410 - val_accuracy: 0.7578
Epoch 10/100
21/21 [==============================] - 0s 2ms/step - loss: 0.5445 - accuracy: 0.7320 - val_loss: 0.5390 - val_accuracy: 0.7534
Epoch 11/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5414 - accuracy: 0.7365 - val_loss: 0.5458 - val_accuracy: 0.7489
Epoch 12/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5333 - accuracy: 0.7485 - val_loss: 0.5214 - val_accuracy: 0.7578
Epoch 13/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5252 - accuracy: 0.7380 - val_loss: 0.5134 - val_accuracy: 0.7578
Epoch 14/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5303 - accuracy: 0.7395 - val_loss: 0.5174 - val_accuracy: 0.7578
Epoch 15/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5338 - accuracy: 0.7440 - val_loss: 0.5353 - val_accuracy: 0.7848
Epoch 16/100
21/21 [==============================] - 0s 2ms/step - loss: 0.5277 - accuracy: 0.7425 - val_loss: 0.4985 - val_accuracy: 0.7848
Epoch 17/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5042 - accuracy: 0.7560 - val_loss: 0.4934 - val_accuracy: 0.7758
Epoch 18/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5019 - accuracy: 0.7545 - val_loss: 0.4873 - val_accuracy: 0.7803
Epoch 19/100
21/21 [==============================] - 0s 3ms/step - loss: 0.5011 - accuracy: 0.7485 - val_loss: 0.4891 - val_accuracy: 0.7892
Epoch 20/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4956 - accuracy: 0.7680 - val_loss: 0.4874 - val_accuracy: 0.7892
Epoch 21/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4991 - accuracy: 0.7605 - val_loss: 0.4780 - val_accuracy: 0.7937
Epoch 22/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4968 - accuracy: 0.7650 - val_loss: 0.4724 - val_accuracy: 0.7892
Epoch 23/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4954 - accuracy: 0.7740 - val_loss: 0.4709 - val_accuracy: 0.7848
Epoch 24/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4917 - accuracy: 0.7769 - val_loss: 0.4763 - val_accuracy: 0.7982
Epoch 25/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4850 - accuracy: 0.7754 - val_loss: 0.4722 - val_accuracy: 0.7848
Epoch 26/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4877 - accuracy: 0.7769 - val_loss: 0.4858 - val_accuracy: 0.7803
Epoch 27/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4749 - accuracy: 0.7814 - val_loss: 0.4665 - val_accuracy: 0.7803
Epoch 28/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4752 - accuracy: 0.7859 - val_loss: 0.4568 - val_accuracy: 0.8027
Epoch 29/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4745 - accuracy: 0.7889 - val_loss: 0.4597 - val_accuracy: 0.8072
Epoch 30/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4827 - accuracy: 0.7979 - val_loss: 0.4534 - val_accuracy: 0.7982
Epoch 31/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4667 - accuracy: 0.8009 - val_loss: 0.4478 - val_accuracy: 0.8117
Epoch 32/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4643 - accuracy: 0.8054 - val_loss: 0.4485 - val_accuracy: 0.8072
Epoch 33/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4644 - accuracy: 0.7814 - val_loss: 0.4516 - val_accuracy: 0.8161
Epoch 34/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4690 - accuracy: 0.8024 - val_loss: 0.4526 - val_accuracy: 0.7848
Epoch 35/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4603 - accuracy: 0.7979 - val_loss: 0.4487 - val_accuracy: 0.8027
Epoch 36/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4599 - accuracy: 0.8129 - val_loss: 0.4398 - val_accuracy: 0.8072
Epoch 37/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4576 - accuracy: 0.8084 - val_loss: 0.4398 - val_accuracy: 0.8072
Epoch 38/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4628 - accuracy: 0.7949 - val_loss: 0.5009 - val_accuracy: 0.7668
Epoch 39/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4810 - accuracy: 0.7829 - val_loss: 0.4420 - val_accuracy: 0.7848
Epoch 40/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4685 - accuracy: 0.8084 - val_loss: 0.4568 - val_accuracy: 0.8027
Epoch 41/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4620 - accuracy: 0.8084 - val_loss: 0.4345 - val_accuracy: 0.8161
Epoch 42/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4592 - accuracy: 0.8144 - val_loss: 0.4353 - val_accuracy: 0.8161
Epoch 43/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4545 - accuracy: 0.8084 - val_loss: 0.4323 - val_accuracy: 0.8072
Epoch 44/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4529 - accuracy: 0.8039 - val_loss: 0.4373 - val_accuracy: 0.8296
Epoch 45/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4536 - accuracy: 0.8114 - val_loss: 0.4491 - val_accuracy: 0.8072
Epoch 46/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4521 - accuracy: 0.8099 - val_loss: 0.4423 - val_accuracy: 0.8072
Epoch 47/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4523 - accuracy: 0.8084 - val_loss: 0.4272 - val_accuracy: 0.8206
Epoch 48/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4495 - accuracy: 0.8024 - val_loss: 0.4273 - val_accuracy: 0.8072
Epoch 49/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4474 - accuracy: 0.8039 - val_loss: 0.4334 - val_accuracy: 0.8161
Epoch 50/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4454 - accuracy: 0.8159 - val_loss: 0.4256 - val_accuracy: 0.8161
Epoch 51/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4509 - accuracy: 0.7979 - val_loss: 0.4252 - val_accuracy: 0.8117
Epoch 52/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4458 - accuracy: 0.8084 - val_loss: 0.4246 - val_accuracy: 0.8251
Epoch 53/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4493 - accuracy: 0.8039 - val_loss: 0.4383 - val_accuracy: 0.8206
Epoch 54/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4650 - accuracy: 0.8054 - val_loss: 0.4343 - val_accuracy: 0.8027
Epoch 55/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4438 - accuracy: 0.8144 - val_loss: 0.4248 - val_accuracy: 0.8027
Epoch 56/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4470 - accuracy: 0.8039 - val_loss: 0.4227 - val_accuracy: 0.8251
Epoch 57/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4382 - accuracy: 0.8129 - val_loss: 0.4213 - val_accuracy: 0.8117
Epoch 58/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4504 - accuracy: 0.8084 - val_loss: 0.4526 - val_accuracy: 0.8072
Epoch 59/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4440 - accuracy: 0.8144 - val_loss: 0.4218 - val_accuracy: 0.8117
Epoch 60/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4368 - accuracy: 0.8144 - val_loss: 0.4205 - val_accuracy: 0.8161
Epoch 61/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4389 - accuracy: 0.8114 - val_loss: 0.4254 - val_accuracy: 0.8206
Epoch 62/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4456 - accuracy: 0.8024 - val_loss: 0.4260 - val_accuracy: 0.8117
Epoch 63/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4355 - accuracy: 0.8054 - val_loss: 0.4252 - val_accuracy: 0.8117
Epoch 64/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4452 - accuracy: 0.8069 - val_loss: 0.4176 - val_accuracy: 0.8161
Epoch 65/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4422 - accuracy: 0.8039 - val_loss: 0.4339 - val_accuracy: 0.7982
Epoch 66/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4439 - accuracy: 0.8114 - val_loss: 0.4196 - val_accuracy: 0.8206
Epoch 67/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4365 - accuracy: 0.8084 - val_loss: 0.4163 - val_accuracy: 0.8117
Epoch 68/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4363 - accuracy: 0.8174 - val_loss: 0.4181 - val_accuracy: 0.8206
Epoch 69/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4376 - accuracy: 0.8144 - val_loss: 0.4215 - val_accuracy: 0.8072
Epoch 70/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4394 - accuracy: 0.8054 - val_loss: 0.4212 - val_accuracy: 0.8161
Epoch 71/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4576 - accuracy: 0.8009 - val_loss: 0.4357 - val_accuracy: 0.8117
Epoch 72/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4391 - accuracy: 0.8084 - val_loss: 0.4303 - val_accuracy: 0.8072
Epoch 73/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4414 - accuracy: 0.8084 - val_loss: 0.4231 - val_accuracy: 0.8072
Epoch 74/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4336 - accuracy: 0.8084 - val_loss: 0.4163 - val_accuracy: 0.8161
Epoch 75/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4339 - accuracy: 0.8114 - val_loss: 0.4218 - val_accuracy: 0.8117
Epoch 76/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4319 - accuracy: 0.8069 - val_loss: 0.4172 - val_accuracy: 0.8206
Epoch 77/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4291 - accuracy: 0.8129 - val_loss: 0.4217 - val_accuracy: 0.8206
Epoch 78/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4302 - accuracy: 0.8174 - val_loss: 0.4181 - val_accuracy: 0.8206
Epoch 79/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4461 - accuracy: 0.8039 - val_loss: 0.4153 - val_accuracy: 0.8027
Epoch 80/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4365 - accuracy: 0.8099 - val_loss: 0.4148 - val_accuracy: 0.8206
Epoch 81/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4441 - accuracy: 0.8114 - val_loss: 0.4301 - val_accuracy: 0.8296
Epoch 82/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4331 - accuracy: 0.8293 - val_loss: 0.4185 - val_accuracy: 0.8296
Epoch 83/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4390 - accuracy: 0.8189 - val_loss: 0.4204 - val_accuracy: 0.8296
Epoch 84/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4359 - accuracy: 0.8174 - val_loss: 0.4160 - val_accuracy: 0.8161
Epoch 85/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4403 - accuracy: 0.7979 - val_loss: 0.4161 - val_accuracy: 0.8117
Epoch 86/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4306 - accuracy: 0.8204 - val_loss: 0.4196 - val_accuracy: 0.8251
Epoch 87/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4636 - accuracy: 0.7994 - val_loss: 0.4232 - val_accuracy: 0.8117
Epoch 88/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4229 - accuracy: 0.8174 - val_loss: 0.4236 - val_accuracy: 0.8296
Epoch 89/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4289 - accuracy: 0.8009 - val_loss: 0.4189 - val_accuracy: 0.8251
Epoch 90/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4310 - accuracy: 0.8249 - val_loss: 0.4196 - val_accuracy: 0.8072
Epoch 91/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4355 - accuracy: 0.8174 - val_loss: 0.4138 - val_accuracy: 0.8117
Epoch 92/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4285 - accuracy: 0.8099 - val_loss: 0.4148 - val_accuracy: 0.8206
Epoch 93/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4301 - accuracy: 0.8278 - val_loss: 0.4130 - val_accuracy: 0.8117
Epoch 94/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4298 - accuracy: 0.8084 - val_loss: 0.4219 - val_accuracy: 0.7937
Epoch 95/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4277 - accuracy: 0.7979 - val_loss: 0.4513 - val_accuracy: 0.8027
Epoch 96/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4427 - accuracy: 0.7934 - val_loss: 0.4192 - val_accuracy: 0.8072
Epoch 97/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4289 - accuracy: 0.8099 - val_loss: 0.4151 - val_accuracy: 0.8161
Epoch 98/100
21/21 [==============================] - 0s 2ms/step - loss: 0.4324 - accuracy: 0.8084 - val_loss: 0.4128 - val_accuracy: 0.8117
Epoch 99/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4355 - accuracy: 0.8009 - val_loss: 0.4285 - val_accuracy: 0.8072
Epoch 100/100
21/21 [==============================] - 0s 3ms/step - loss: 0.4321 - accuracy: 0.8144 - val_loss: 0.4166 - val_accuracy: 0.8072

plt.plot(history.history['accuracy'])
plt.plot(history.history['val_accuracy'])
plt.title('Precisión del modelo')
plt.xlabel('Epoch')
plt.ylabel('Presición')
plt.legend(['Train', 'Test'])
plt.show()

scores = model.evaluate(X_test, y_test)
print('%s: %.4f%%' % (model.metrics_names[1], scores[1] * 100))

7/7 [==============================] - 0s 1ms/step - loss: 0.4166 - accuracy: 0.8072
accuracy: 80.7175%