from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Dense
 
model = Sequential()

# 신경망 모델 구조 정의
model.add(Dense(3, input_dim=2, activation='sigmoid'))
model.add(Dense(1, activation='sigmoid'))

# 컴파일 단계, 옵티마이저와 손실함수, 측정지표를 연결해서 계산 그래프를 구성을 마무리 합니다.
model.compile(optimizer='sgd', loss='mse', metrics=['mae', 'mse'])
# model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy']) # 분류인 경우 예시

results = model.fit(X,y, epochs=50)
results.history.keys()

plt.plot(results.history['loss'])

Pred_Test

from tensorflow.keras.preprocessing.image import load_img
from tensorflow.keras.preprocessing.image import img_to_array
import matplotlib.pyplot as plt
import numpy as np

img_height = 256
img_width = 256

filename = ['./data/test/cat.jpg','./data/test/dog.jpg','./data/test/panda.jpg']

predict_dictionary = {0:'cat',1:'dog',2:'panda'}

for file in filename:
    original = load_img(file, target_size = (img_height,img_width))

    numpy_image = img_to_array(original)
    plt.imshow(np.uint8(numpy_image))
    plt.show()
    print("numpy array size : ", numpy_image.shape)
    image_batch = np.expand_dims(numpy_image , axis = 0)

    predict = np.argmax(model.predict(image_batch/255.))

    print('결과 : ',predict_dictionary[predict])