典型的卷积神经网络。
- Keras傻瓜式读取数据：自动下载，自动解压，自动加载。
- # X_train:
array([[[[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
...,
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.]]],
...,
[[[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
...,
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.],
[ 0., 0., 0., ..., 0., 0., 0.]]]], dtype=float32)
- # y_train:
array([5, 0, 4, ..., 5, 6, 8], dtype=uint8)
但需要二值化作为output：np_utils.to_categorical(y_train, nb_classes)
- # Y_train:
Y_train[0] Out[56]: array([ 0., 0., 0., 0., 0., 1., 0., 0., 0., 0.]) Y_train[1] Out[57]: array([ 1., 0., 0., 0., 0., 0., 0., 0., 0., 0.]) Y_train[2] Out[58]: array([ 0., 0., 0., 0., 1., 0., 0., 0., 0., 0.])
Code:
#coding:utf-8
import os
from PIL import Image
import numpy as np
#读取文件夹mnist下的42000张图片，图片为灰度图，所以为1通道，
#如果是将彩色图作为输入,则将1替换为3，并且data[i,:,:,:] = arr改为data[i,:,:,:] = [arr[:,:,0],arr[:,:,1],arr[:,:,2]]
def load_data():
data = np.empty((42000,1,28,28),dtype="float32")
label = np.empty((42000,),dtype="uint8")
imgs = os.listdir("./mnist")
num = len(imgs)
for i in range(num):
img = Image.open("./mnist/"+imgs[i])
arr = np.asarray(img,dtype="float32")
data[i,:,:,:] = arr
label[i] = int(imgs[i].split(‘.‘)[0])
return data,label
读取原始图片
Code: a Multilayer Perceptron
import numpy as np np.random.seed(1337) # for reproducibility import os from keras.datasets import mnist #自动下载 # import 套路 from keras.models import Sequential from keras.layers.core import Dense, Dropout, Activation from keras.optimizers import RMSprop from keras.utils import np_utils batch_size = 128 #Number of images used in each optimization step nb_classes = 10 #One class per digit nb_epoch = 12 #Number of times the whole data is used to learn (X_train, y_train), (X_test, y_test) = mnist.load_data() #Flatten the data, MLP doesn‘t use the 2D structure of the data. 784 = 28*28 X_train = X_train.reshape(60000, 784) X_test = X_test.reshape(10000, 784) #Make the value floats in [0;1] instead of int in [0;255] --> [归一化] X_train = X_train.astype(‘float32‘) X_test = X_test.astype(‘float32‘) X_train /= 255 X_test /= 255 #Display the shapes to check if everything‘s ok print(X_train.shape[0], ‘train samples‘) print(X_test.shape[0], ‘test samples‘) # convert class vectors to binary class matrices (ie one-hot vectors)Y_train = np_utils.to_categorical(y_train, nb_classes) Y_test = np_utils.to_categorical(y_test, nb_classes) #Define the model achitecture model = Sequential()######################################################################################## model.add(Dense(512, input_shape=(784,))) model.add(Activation(‘relu‘)) model.add(Dropout(0.2)) model.add(Dense(512)) model.add(Activation(‘relu‘)) model.add(Dropout(0.2)) model.add(Dense(10)) #Last layer with one output per class model.add(Activation(‘softmax‘)) #We want a score simlar to a probability for each class ######################################################################################## #Use rmsprop to do the gradient descent see http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf #and http://cs231n.github.io/neural-networks-3/#ada rms = RMSprop() #The function to optimize is the cross entropy between the true label and the output (softmax) of the model model.compile(loss=‘categorical_crossentropy‘, optimizer=rms, metrics=["accuracy"]) #Make the model learn --> [Training] model.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch, verbose=2, validation_data=(X_test, Y_test)) #Evaluate how the model does on the test set score = model.evaluate(X_test, Y_test, verbose=0) print(‘Test score:‘, score[0]) print(‘Test accuracy:‘, score[1])
Code: a Convolutional Neural Network
import numpy as np np.random.seed(1337) # for reproducibility import os from keras.datasets import mnist from keras.models import Sequential from keras.layers.core import Dense, Dropout, Activation, Flatten from keras.layers.convolutional import Convolution2D, MaxPooling2D from keras.utils import np_utils batch_size = 128 nb_classes = 10 nb_epoch = 12 # input image dimensions img_rows, img_cols = 28, 28 # number of convolutional filters to use nb_filters = 32 # size of pooling area for max pooling nb_pool = 2 # convolution kernel size nb_conv = 3 # the data, shuffled and split between train and test sets (X_train, y_train), (X_test, y_test) = mnist.load_data() #Add the depth in the input. Only grayscale so depth is only one #see http://cs231n.github.io/convolutional-networks/#overview 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) #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 /= 255 X_test /= 255 #Display the shapes to check if everything‘s ok 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 (ie one-hot vectors) Y_train = np_utils.to_categorical(y_train, nb_classes) Y_test = np_utils.to_categorical(y_test, nb_classes) ############################################################################################## 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(nb_filters, nb_conv, nb_conv, border_mode=‘valid‘, input_shape=(1, img_rows, img_cols)))model.add(Activation(‘relu‘)) model.add(Convolution2D(nb_filters, nb_conv, nb_conv)) model.add(Activation(‘relu‘)) #For an explanation on pooling layers see http://cs231n.github.io/convolutional-networks/#pool model.add(MaxPooling2D(pool_size=(nb_pool, nb_pool))) model.add(Dropout(0.25)) #Flatten the 3D output to 1D tensor for a fully connected layer to accept the input model.add(Flatten()) model.add(Dense(128)) model.add(Activation(‘relu‘)) model.add(Dropout(0.5)) model.add(Dense(nb_classes)) #Last layer with one output per class model.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/#ada model.compile(loss=‘categorical_crossentropy‘, optimizer=‘adadelta‘, metrics=["accuracy"]) #Make the model learn model.fit(X_train, Y_train, batch_size=batch_size, nb_epoch=nb_epoch, verbose=1, validation_data=(X_test, Y_test)) #Evaluate how the model does on the test set score = model.evaluate(X_test, Y_test, verbose=0) print(‘Test score:‘, score[0]) print(‘Test accuracy:‘, score[1])
另一个卷积示例：
#coding:utf-8
‘‘‘
GPU run command:
THEANO_FLAGS=mode=FAST_RUN,device=gpu,floatX=float32 python cnn.py
CPU run command:
python cnn.py
‘‘‘
#导入各种用到的模块组件
from __future__ import absolute_import
from __future__ import print_function
from keras.preprocessing.image import ImageDataGenerator
from keras.models import Sequential
from keras.layers.core import Dense, Dropout, Activation, Flatten
from keras.layers.advanced_activations import PReLU
from keras.layers.convolutional import Convolution2D, MaxPooling2D
from keras.optimizers import SGD, Adadelta, Adagrad
from keras.utils import np_utils, generic_utils
from six.moves import range
from data import load_data
import random
import numpy as np
np.random.seed(1024) # for reproducibility
#加载数据
data, label = load_data()
#打乱数据
index = [i for i in range(len(data))]
random.shuffle(index)
data = data[index]
label = label[index]
print(data.shape[0], ‘ samples‘)
#label为0~9共10个类别，keras要求格式为binary class matrices,转化一下，直接调用keras提供的这个函数
label = np_utils.to_categorical(label, 10)
###############
#开始建立CNN模型
###############
#生成一个model
model = Sequential()
#【第一个卷积层】，4个卷积核，每个卷积核大小5*5。1表示输入的图片的通道,灰度图为1通道。
#border_mode可以是valid或者full，参见这里：http://blog.csdn.net/niuwei22007/article/details/49366745
#激活函数用tanh
#你还可以在model.add(Activation(‘tanh‘))后加上dropout的技巧: model.add(Dropout(0.5))
model.add(Convolution2D(4, 5, 5, border_mode=‘valid‘,input_shape=(1,28,28)))
model.add(Activation(‘tanh‘))
#【第二个卷积层】，8个卷积核，每个卷积核大小3*3。4表示输入的特征图个数，等于上一层的卷积核个数
#激活函数用tanh
#采用maxpooling，poolsize为(2,2)
model.add(Convolution2D(8, 3, 3, border_mode=‘valid‘))
model.add(Activation(‘tanh‘))
model.add(MaxPooling2D(pool_size=(2, 2)))
#【第三个卷积层】，16个卷积核，每个卷积核大小3*3
#激活函数用tanh
#采用maxpooling，poolsize为(2,2)
model.add(Convolution2D(16, 3, 3, border_mode=‘valid‘))
model.add(Activation(‘relu‘))
model.add(MaxPooling2D(pool_size=(2, 2)))
#【全连接层】，先将前一层输出的二维特征图flatten为一维的。
#Dense就是隐藏层。16就是上一层输出的特征图个数。4是根据每个卷积层计算出来的：(28-5+1)得到24,(24-3+1)/2得到11，(11-3+1)/2得到4
#全连接有128个神经元节点,初始化方式为normal
model.add(Flatten())
model.add(Dense(128, init=‘normal‘))
model.add(Activation(‘tanh‘))
#【Softmax分类】，输出是10类别
model.add(Dense(10, init=‘normal‘))
model.add(Activation(‘softmax‘))
#############
#开始训练模型
##############
#使用SGD + momentum
#model.compile里的参数loss就是损失函数(目标函数)
sgd = SGD(lr=0.05, decay=1e-6, momentum=0.9, nesterov=True)
model.compile(loss=‘categorical_crossentropy‘, optimizer=sgd,metrics=["accuracy"])
#调用fit方法，就是一个训练过程. 训练的epoch数设为10，batch_size为100．
#数据经过随机打乱shuffle=True。verbose=1，训练过程中输出的信息，0、1、2三种方式都可以，无关紧要。show_accuracy=True，训练时每一个epoch都输出accuracy。
#validation_split=0.2，将20%的数据作为验证集。
model.fit(data, label, batch_size=100, nb_epoch=10,shuffle=True,verbose=1,validation_split=0.2)
时间： 2024-08-01 10:46:03