数字图像 深度学习实验

1. 根据mnist_cnn_to_student.py代码的提示，编写卷积神经网络代码，并尝试设计AlexNet框架（包含5个卷积层的神经网络），用于手写数字识别。

2.完成情感分析代码（基于循环神经网络）的运行与调试。

1.卷积神经网络

'''Trains a simple convnet on the MNIST dataset.

Gets to 99.25% test accuracy after 12 epochs

(there is still a lot of margin for parameter tuning).

16 seconds per epoch on a GRID K520 GPU.

'''

from __future__ import print_function

import keras

from keras.datasets import mnist

from keras.models import Sequential

from keras.layers import Dense, Dropout, Flatten

from keras.layers import Conv2D, MaxPooling2D

from keras import backend as K

batch_size = 128   # 一次训练所选取的样本数

num_classes = 10   # 分类的类别数

epochs = 12        # 完整的数据集所训练的次数

# 训练集有1000个样本，batchsize=10，那么：训练完整个样本集需要： 100次iteration，1次epoch。

# input image dimensions  数据的维数

img_rows, img_cols = 28, 28

# the data, split between train and test sets

(x_train, y_train), (x_test, y_test) = mnist.load_data()

if K.image_data_format() == 'channels_first':

    x_train = x_train.reshape(x_train.shape[0], 1, img_rows, img_cols)

    x_test = x_test.reshape(x_test.shape[0], 1, img_rows, img_cols)

    input_shape = (1, img_rows, img_cols)

else:

    x_train = x_train.reshape(x_train.shape[0], img_rows, img_cols, 1)

    x_test = x_test.reshape(x_test.shape[0], img_rows, img_cols, 1)

    input_shape = (img_rows, img_cols, 1)

x_train = x_train.astype('float32')   # 将uint8型数据转成float32

x_test = x_test.astype('float32')

x_train /= 255                     # 归一化处理

x_test /= 255

print('x_train shape:', x_train.shape)

print(x_train.shape[0], 'train samples')

print(x_test.shape[0], 'test samples')

# convert class vectors to binary class matrices  将类别向量转换为二进制类矩阵

y_train = keras.utils.to_categorical(y_train, num_classes)

y_test = keras.utils.to_categorical(y_test, num_classes)

model = Sequential()   # 接下来定义顺序模型，输入 28*28大小的图像

# 卷积层1，首先加入2D卷积层，卷积核大小3*3，通道数32，激活函数为relu