我对机器学习还比较陌生，所以我在尝试一些示例代码。代码中指定的图像尺寸是（28,28），但不知为何我总是得到相同的ValueError错误，我搞不清楚为什么会这样。
这是我的代码：
import pandas as pdimport numpy as npnp.random.seed(1337) # for reproducibilityfrom keras.models import Sequentialfrom keras.layers.core import Dense, Dropout, Activation, Flattenfrom keras.layers.convolutional import Convolution2D, MaxPooling2Dfrom keras.utils import np_utils# input image dimensionsimg_rows, img_cols = 28, 28batch_size = 128 # Number of images used in each optimization stepnb_classes = 10 # One class per digitnb_epoch = 35 # Number of times the whole data is used to learn# Read the train and test datasetstrain = pd.read_csv("../input/train.csv").valuestest = pd.read_csv("../input/test.csv").values# Reshape the data to be used by a Theano CNN. Shape is# (nb_of_samples, nb_of_color_channels, img_width, img_heigh)X_train = train[:, 1:].reshape(train.shape[0], 1, img_rows, img_cols)X_test = test.reshape(test.shape[0], 1, img_rows, img_cols)y_train = train[:, 0] # First data is label (already removed from X_train)# Make the value floats in [0;1] instead of int in [0;255]X_train = X_train.astype('float32')X_test = X_test.astype('float32')X_train /= 255X_test /= 255# convert class vectors to binary class matrices (ie one-hot vectors)Y_train = np_utils.to_categorical(y_train, nb_classes)#Display the shapes to check if everything's okprint('X_train shape:', X_train.shape)print('Y_train shape:', Y_train.shape)print('X_test shape:', X_test.shape)model = Sequential()# For an explanation on conv layers see http://cs231n.github.io/convolutional-networks/#conv# By default the stride/subsample is 1# border_mode "valid" means no zero-padding.# If you want zero-padding add a ZeroPadding layer or, if stride is 1 use border_mode="same"model.add(Convolution2D(12, 5, 5, border_mode='valid',input_shape=(1,img_rows, img_cols)))model.add(Activation('relu'))# For an explanation on pooling layers see http://cs231n.github.io/convolutional-networks/#poolmodel.add(MaxPooling2D(pool_size=(2, 2)))model.add(Dropout(0.15))model.add(Convolution2D(24, 5, 5))model.add(Activation('relu'))model.add(MaxPooling2D(pool_size=(2, 2)))model.add(Dropout(0.15))# Flatten the 3D output to 1D tensor for a fully connected layer to accept the inputmodel.add(Flatten())model.add(Dense(180))model.add(Activation('relu'))model.add(Dropout(0.5))model.add(Dense(100))model.add(Activation('relu'))model.add(Dropout(0.5))model.add(Dense(nb_classes)) #Last layer with one output per classmodel.add(Activation('softmax')) #We want a score simlar to a probability for each class# The function to optimize is the cross entropy between the true label and the output (softmax) of the model# We will use adadelta to do the gradient descent see http://cs231n.github.io/neural-networks-3/#adamodel.compile(loss='categorical_crossentropy', optimizer='adadelta', metrics=["accuracy"])# Make the model learnmodel.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch, verbose=1)# Predict the label for X_testyPred = model.predict_classes(X_test)# Save prediction in file for Kaggle submissionnp.savetxt('mnist-pred.csv', np.c_[range(1,len(yPred)+1),yPred], delimiter=',', header = 'ImageId,Label', comments = '', fmt='%d')回答：
所以问题出在所使用的卷积尺寸上。卷积操作通常会减小图像的尺寸。同样，每次池化操作也会减小尺寸。你使用的是非常小的图像，但却应用了为更大图像设计的模型架构，因此在某个点上，经过一次卷积/池化后，输出的图像尺寸实际上比后续的滤波器尺寸还要小，这是一个未定义的操作。
要暂时解决这个问题，可以移除第二个卷积和最大池化层，因为这些操作（使用提供的参数）无法在如此小的数据上进行。总的来说，你应该首先理解卷积是如何工作的，而不是直接应用别人的模型，因为参数对性能至关重要——如果你应用的变换过度降低了分辨率，你将无法学到任何东西。因此，一旦你对卷积的工作原理有了直觉，你可以返回并尝试不同的架构，但没有一个“神奇”的公式可以确定架构，因此我无法提供“直接有效”的参数——先从移除这个额外的卷积和池化开始，然后在你对数据和模型有了更好的理解后，再尝试其他可能性。
