[Deep Learning] Keras 기초

설명

• y = x 방정식 학습 예제

기본 코드

import numpy as np
from keras.models import Sequential
from keras.layers import Dense

x = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
y = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])

model = Sequential()
model.add(Dense(1, input_dim=1, activation='relu'))
model.compile(loss='mean_squared_error', optimizer='adam', metrics=['accuracy'])
model.fit(x, y, epochs=500, batch_size=1)

loss, acc = model.evaluate(x, y, batch_size=1)

print("loss : ", loss)
print("acc : ", acc)
print(model.predict(x))

loss :  0.0
acc :  0.10000000149011612
[[ 1.]
 [ 2.]
 [ 3.]
 [ 4.]
 [ 5.]
 [ 6.]
 [ 7.]
 [ 8.]
 [ 9.]
 [10.]]

학습 데이터와 테스트 데이터를 나눠 학습 코드

import numpy as np
from keras.layers import Dense
from keras.models import Sequential

x_train = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
y_train = np.array([1, 2, 3, 4, 5, 6, 7, 8, 9, 10])
x_test = np.array([101, 102, 103, 104, 105, 106, 107, 108, 109, 110])
y_test = np.array([101, 102, 103, 104, 105, 106, 107, 108, 109, 110])

model = Sequential()
model.add(Dense(5, input_dim=1, activation='relu'))
model.add(Dense(3))
model.add(Dense(1, activation='relu'))
model.summary()
model.compile(loss='mse', optimizer='adam', metrics=['accuracy'])
model.fit(x_train, y_train, epochs=300, batch_size=1, validation_data=(x_test, y_test))

loss, acc = model.evaluate(x_test, y_test, batch_size=1)

print("loss : ", loss)
print("acc : ", acc)
print(model.predict(x_test))

loss :  0.8712489008903503
acc :  0.0
[[101.89025 ]
 [102.899734]
 [103.90925 ]
 [104.91874 ]
 [105.92826 ]
 [106.93776 ]
 [107.947266]
 [108.95677 ]
 [109.966286]
 [110.97578 ]]

Train, Val, Test 데이터를 준비하여 학습

import numpy as np
from keras.models import Sequential
from keras.layers import Dense
from sklearn.model_selection import train_test_split

x = np.array(range(1, 101))
y = np.array(range(1, 101))

# Train:Val:Test 데이터 비율을 6:2:2 비율로 세팅
x_train, x_test, y_train, y_test = train_test_split(x, y, random_state=66, test_size=0.4, shuffle=False)
x_val, x_test, y_val, y_test = train_test_split(x_test, y_test, random_state=66, test_size=0.5, shuffle=False)

model = Sequential()
# model.add(Dense(5, input_dim=1, activation="relu"))
model.add(Dense(5, input_shape=(1,), activation="relu"))
model.add(Dense(3))
model.add(Dense(4))
model.add(Dense(1))
model.compile(loss="mse", optimizer="adam", metrics=["mse"])
model.fit(x_train, y_train, epochs=300, batch_size=1, validation_data=(x_val, y_val))

mse = model.evaluate(x_test, y_test, batch_size=1)
print("mse : ", mse)

y_predict = model.predict(x_test)
print("y_predict : ", y_predict)

mse :  [6.411573938436277e-09, 6.411573938436277e-09]
y_predict :  [[ 81.000084]
 [ 82.00007 ]
 [ 83.00008 ]
 [ 84.00008 ]
 [ 85.00008 ]
 [ 86.00008 ]
 [ 87.000084]
 [ 88.000084]
 [ 89.00008 ]
 [ 90.00008 ]
 [ 91.000084]
 [ 92.000084]
 [ 93.000084]
 [ 94.00008 ]
 [ 95.00008 ]
 [ 96.00008 ]
 [ 97.00009 ]
 [ 98.00009 ]
 [ 99.00009 ]
 [100.00008 ]]

1입력 1출력 예제

import numpy as np
from keras.layers import Dense, Input
from keras.models import Model
from sklearn.model_selection import train_test_split

x = np.array(range(1, 101))
y = np.array(range(1, 101))

# Train:Val:Test 데이터 비율을 6:2:2 비율로 세팅
x_train, x_test, y_train, y_test = train_test_split(x, y, random_state=66, test_size=0.4, shuffle=False)
x_val, x_test, y_val, y_test = train_test_split(x_test, y_test, random_state=66, test_size=0.5, shuffle=False)

input1 = Input(shape=(1,))  # 1 입력
dense1 = Dense(5, activation='relu')(input1)
dense2 = Dense(3)(dense1)
dense3 = Dense(4)(dense2)
output1 = Dense(1)(dense3)  # 1 출력

model = Model(inputs=input1, outputs=output1)
# model.summary()

model.compile(loss="mse", optimizer="adam", metrics=["mse"])
model.fit(x_train, y_train, epochs=100, batch_size=1, validation_data=(x_val, y_val))

mse = model.evaluate(x_test, y_test, batch_size=1)
print("mse : ", mse)

y_predict = model.predict(x_test)
print("y_predict : ", y_predict)

mse :  [1.65891828052267e-10, 1.65891828052267e-10]
y_predict :  [[ 81.      ]
 [ 82.      ]
 [ 83.      ]
 [ 84.      ]
 [ 85.000015]
 [ 86.00001 ]
 [ 87.      ]
 [ 88.00001 ]
 [ 89.000015]
 [ 90.000015]
 [ 91.000015]
 [ 92.000015]
 [ 93.000015]
 [ 94.000015]
 [ 95.00001 ]
 [ 96.00001 ]
 [ 97.      ]
 [ 98.00003 ]
 [ 99.000015]
 [100.      ]]

N입력 N출력 예제

import numpy as np
from keras.layers import Dense
from keras.models import Sequential
from sklearn.model_selection import train_test_split

x = np.array([range(100), range(301, 401)])
y = np.array([range(100), range(301, 401)])

x = np.transpose(x)
y = np.transpose(y)

# Train:Val:Test 데이터 비율을 6:2:2 비율로 세팅
x_train, x_test, y_train, y_test = train_test_split(x, y, random_state=66, test_size=0.4, shuffle=False)
x_val, x_test, y_val, y_test = train_test_split(x_test, y_test, random_state=66, test_size=0.5, shuffle=False)

model = Sequential()
# model.add(Dense(5, input_dim=2, activation="relu"))
model.add(Dense(5, input_shape=(2,), activation="relu"))  # 2 입력
model.add(Dense(3))
model.add(Dense(4))
model.add(Dense(2))  # 2 출력
model.compile(loss="mse", optimizer="adam", metrics=["mse"])
model.fit(x_train, y_train, epochs=300, batch_size=1, validation_data=(x_val, y_val))

mse = model.evaluate(x_test, y_test, batch_size=1)
print("mse : ", mse)

y_predict = model.predict(x_test)
print("y_predict : ", y_predict)

mse :  [1.9659782424241712e-07, 1.9659782424241712e-07]
y_predict :  [[ 80.0005   380.99988 ]
 [ 81.0005   381.99985 ]
 [ 82.0005   382.99985 ]
 [ 83.00054  383.99985 ]
 [ 84.000534 384.99985 ]
 [ 85.00054  385.99985 ]
 [ 86.000565 386.99985 ]
 [ 87.00058  387.99985 ]
 [ 88.0006   388.99985 ]
 [ 89.00059  389.99985 ]
 [ 90.00057  390.99982 ]
 [ 91.000626 391.99985 ]
 [ 92.00065  392.99982 ]
 [ 93.00065  393.99982 ]
 [ 94.00063  394.9998  ]
 [ 95.00063  395.9998  ]
 [ 96.00065  396.9998  ]
 [ 97.000694 397.9998  ]
 [ 98.00071  398.9998  ]
 [ 99.000725 399.9998  ]]