【发布时间】:2018-12-14 14:09:06
【问题描述】:
我编写了一个用于二进制分类的 CNN 模型。我的数据集有偏差(56000 张图像的第 1 类和 3000 张图像的第 2 类)。 我正在测试 108 张图片(每类 54 张)。我的模型将每个图像都预测为第 1 类。你能告诉我我的模型有什么问题吗?我该如何改进它?
IMG_SIZE = 32
LR = 1e-1
convnet = input_data(shape=[None, IMG_SIZE, IMG_SIZE, 3], name='input')
convnet = conv_2d(convnet, 32, 5, activation='relu')
convnet = max_pool_2d(convnet, 5)
convnet = conv_2d(convnet, 64, 5, activation='relu')
convnet = max_pool_2d(convnet, 5)
convnet = conv_2d(convnet, 128, 5, activation='relu')
convnet = max_pool_2d(convnet, 5)
convnet = conv_2d(convnet, 64, 5, activation='relu')
convnet = max_pool_2d(convnet, 5)
convnet = conv_2d(convnet, 32, 5, activation='relu')
convnet = max_pool_2d(convnet, 5)
convnet = fully_connected(convnet, 1024, activation='relu')
convnet = dropout(convnet, 0.8)
convnet = fully_connected(convnet, 2, activation='softmax')
convnet = regression(convnet, optimizer='adam', learning_rate=LR, loss='categorical_crossentropy', name='targets')
model = tflearn.DNN(convnet, tensorboard_dir='/home/anas/Argentinadata/log')
train = training_data[:50000]
test =training_data[50000:]
X = np.array([i[0] for i in train]).reshape(-1,IMG_SIZE,IMG_SIZE,3)
Y = [i[1] for i in train]
test_x = np.array([i[0] for i in test]).reshape(-1,IMG_SIZE,IMG_SIZE,3)
test_y = [i[1] for i in test]
print(len(test_x))
print(len(X))
print(len(Y))
model.fit({'input': X}, {'targets': Y}, n_epoch=25, validation_set=({'input': test_x}, {'targets': test_y}),
snapshot_step=500, show_metric=True, run_id=MODEL_NAME)
【问题讨论】:
标签: python-3.x tensorflow neural-network conv-neural-network