Question

显微镜图像为.tif格式，并具有以下规格：

颜色模型：R（ed）G（reen）B（lue）
尺寸：2048 x 1536像素
像素刻度：0.42μmx 0.42μm
内存空间：10-20 MB（大约）
标签类型：按图像显示
4类：良性，侵入性，原位，正常

CNN代码：

while carryOn:
    for event in pygame.event.get():
        if event.type==pygame.QUIT:
            carryOn=False
        elif event.type==pygame.KEYDOWN:
            if event.key==pygame.K_x:
                carryOn=False

    keys = pygame.key.get_pressed()
    if keys[pygame.K_a]:
        playerChar.moveLeftP(MOVE)
    if keys[pygame.K_d]:
        playerChar.moveRightP(MOVE)
    if keys[pygame.K_w]:
        playerChar.moveUpP(MOVE)
    if keys[pygame.K_s]:
        playerChar.moveDownP(MOVE)

    current_time = pygame.time.get_ticks()
    if start and current_time > next_zombie_time:
        next_zombie_time = current_time + 2000

        on_screen_rect = pygame.Rect(zombie_rad, zombie_rad, size[0]-2*zombie_rad, size[1]-2*zombie_rad)
        zombie_pos = (-1, -1)
        while not on_screen_rect.collidepoint(zombie_pos):
            dist  = random.randint(*zombie_dist)
            angle = random.random() * math.pi * 2 
            p_pos = (playerChar.rect.centerx, playerChar.rect.centery)
            zombie_pos = (p_pos[0] + dist * math.sin(angle), p_pos[1] + dist * math.cos(angle))

        new_pos = (random.randrange(0, size[0]), random.randrange(0, size[1]))
        new_zombie = Zombie(RED, zombie_pos, zombie_rad, 0)
        zombie_list.append(new_zombie)

    # update all the positions of the zombies
    for zombie in zombie_list:
        zombie.move_towards_Char(playerChar)

    screen.fill(BGColor)
    screen.blit(playerChar.image,playerChar.rect)

    # draw all the zombies
    for zombie in zombie_list:
        zombie.draw()

    pygame.display.flip()
    clock.tick(60)

输出：

from keras.models import Sequential
from keras.layers import Convolution2D
from keras.layers import MaxPooling2D
from keras.layers import Flatten
from keras.layers import Dense


classifier = Sequential()


classifier.add(Convolution2D(32, 3, 3, input_shape = (64, 64, 3), activation = 'relu'))


classifier.add(MaxPooling2D(pool_size = (2, 2)))


classifier.add(Convolution2D(32, 3, 3, activation = 'relu'))
classifier.add(MaxPooling2D(pool_size = (2, 2)))

classifier.add(Flatten())


classifier.add(Dense(output_dim = 128, activation = 'relu'))
classifier.add(Dense(output_dim = 1, activation = 'sigmoid'))


classifier.compile(optimizer = 'adam', loss = 'binary_crossentropy', metrics = ['accuracy'])


from keras.preprocessing.image import ImageDataGenerator

train_datagen = ImageDataGenerator(rescale = 1./255,
                                   shear_range = 0.2,
                                   zoom_range = 0.2,
                                   horizontal_flip = True)

test_datagen = ImageDataGenerator(rescale = 1./255,
                                   shear_range = 0.2,
                                   zoom_range = 0.2,
                                   horizontal_flip = True)

training_set = train_datagen.flow_from_directory('BACH ICIAR 2018/ICIAR2018_BACH_Challenge/Photos',
                                                 target_size = (64, 64),
                                                 batch_size = 1,
                                                 class_mode = 'binary')

test_set = test_datagen.flow_from_directory('BACH ICIAR 2018/ICIAR2018_BACH_Challenge_TestDataset',
                                            target_size = (64, 64),
                                            batch_size = 1,
                                            class_mode = 'binary')

classifier.fit_generator(training_set,
                         samples_per_epoch = 8000,
                         nb_epoch = 1,
                         validation_data = test_set,
                         nb_val_samples = 2000)

如果我更改它，则会出现错误，如下所示：

Found 400 images belonging to 4 classes.
Found 100 images belonging to 1 classes.
331/8000 [>.............................] - ETA: 23:37 - loss: -7.9829 - acc: 0.2538

输出：

classifier.add(Dense(output_dim = 4, activation = 'softmax'))

如果我更改它，则会出现错误，如下所示：

TypeError: softmax() got an unexpected keyword argument 'axis'

输出：

classifier.compile(optimizer = 'adam', loss = 'categorical_crossentropy', metrics = ['accuracy'])

为什么损失值是负数？我需要对代码进行哪些更改？这是什么意思？可能的原因是什么？

谢谢您的回复

CNN-负损失值

0 个答案: