[DL] XOR 학습시키기 2

04 Dec 2020

머신러닝

XOR 학습시키기(TF2.xx)

TF2.xx를 사용한 DNN 모델을 만든다.

Library

사용한 Library를 나열한다.

import numpy as np
import tensorflow as tf
import matplotlib.pyplot as plt

from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
from tensorflow.keras.optimizers import SGD

from sklearn.metrics import classification_report

학습 진행

TF2.xx로 학습을 진행해본다.

XOR DataSet

x_data = np.array([[0,0],[1,0],[0,1],[1,1]], dtype=np.float32)
t_data = np.array([[0],[1],[1],[0]], dtype=np.float32)

model 및 layer 생성, compile

keras_model = Sequential()

keras_model.add(Dense(100, activation='sigmoid', input_size=(x_data.shape[1],)))
## Flatten 없이 Dense로만 표한 가능하다.
keras_model.add(Dense(6, activation='sigmoid'))
keras_model.add(Dense(1, activation='sigmoid'))

keras_model.compile(optimizer=SGD,
                    loss='binary_crossentropy',
                    metrics=['accuracy'])

학습 진행

history =keras_model.fit(x_data,
          	             t_data,
                		 epochs=30000,
	 	                 verbose=0)

결과

predict_val = keras_model.predict(x_data)
print(classfication_report(t_data.ravel(), tf.cast(predict_val>=0.5, dtype=tf.float32 ).numpy().ravel()))
##               precision    recall  f1-score   support
## 
##          0.0       1.00      1.00      1.00         2
##          1.0       1.00      1.00      1.00         2
## 
##     accuracy                           1.00         4
##    macro avg       1.00      1.00      1.00         4
## weighted avg       1.00      1.00      1.00         4

Accuracy 그래프

plt.plot(history.history['accuracy'], color='b')
plt.show()