Recording the process of studying keras
Model
Sequential Model
Buliding a simple Sequential Model
from keras.models import Sequential
model = Sequential()
Add layers for the model
from keras.layers import Dense, Dropout
model.add(Dense(64, input_dim = 20, activation = 'relu'))
model.add(Dropout(0.5))
model.add(Dense(64, activation = 'relu'))
model.add(Dropout(0.5))
model.add(Dense(1, activation = 'sigmoid'))
Compiling our model.
model.compile(loss='binary_crossentropy',
optimizer='rmsprop',
metrics=['accuracy'])
Generate dummy data
x_train = np.random.random((1000, 20))
y_train = np.random.randint(2, size=(1000, 1))
x_test = np.random.random((100, 20))
y_test = np.random.randint(2, size=(100, 1))
Start training
model.fit(x_train, y_train,batch_size = 200,epochs = 10)
Layers
Dense
keras.layers.Dense(units, input_dim, activation)
units output space
input_dim the requirement of the input_dim, only requiring it at the begining
activation activation function
Dropout
keras.layers.Dropout(rate)
rate the rate of nodes which are ignored during each batch.
Conv2D
keras.layers.Conv2D(filters, kernel_size, strides=(1, 1), padding='valid',dilation_rate=(1, 1), activation=None, use_bias=True, kernel_regularizer=None)
filters the number of the filters/ output channels
kernel_size int or tuple
padding ‘valid’ means no padding; ‘same’ means reshape the output with the same size of input
activation activation function
kernel_regularizer Regularizer function applied to the kernel weights matrix
Upsampling2D
keras.layers.UpSampling2D(size=(2, 2), data_format=None)
size int or tuple, the upsampling factors for rows and columns.
PS
tensor
>>> import numpy as np
>>> import math
>>> a = np.array([[1,2,3],[4,5,6]])
>>> print(a)
[[1 2 3]
[4 5 6]]
>>> a.shape()
(2, 3)
>>> b = np.array([1,2,3])
>>> print(b)
[1 2 3]
>>> np.shape(b)
(3,)
与矩阵中概念不同，shape(2,3)指第一维有2个元素，第二维有3个元素
batch_size the number of sample in one iteration.
epochs go through all the samples.
From above， epoch = nums of iterations = nusm of samples / batch_size
Example
MLP (Multi-Layer Perceptron)
除了input layer, output layer，至少有一个hidden layer ，形式上full-connected
Mnist using MLP
import numpy
from keras.datasets import mnist
from keras.models import Sequential
from keras.layers import Dense
from keras.layers import Dropout
from keras.utils import np_utils
seed = 7
numpy.random.seed(seed)
data = numpy.load('mnist.npz')
X_train, y_train = data['x_train'],data['y_train']
X_test, y_test = data['x_test'], data['y_test']
num_pixels = X_train.shape[1] * X_train.shape[2]
X_train = X_train.reshape(X_train.shape[0], num_pixels).astype('float32')
X_test = X_test.reshape(X_test.shape[0], num_pixels).astype('float32')
# for 0~255 to 0~1
X_train = X_train/255
X_test = X_test/255
# convert to 'one-hot' format
y_train = np_utils.to_categorical(y_train)
y_test = np_utils.to_categorical(y_test)
num_classes = y_test.shape[1]
# create model
model = Sequential()
model.add(Dense(num_pixels, input_dim = num_pixels, activation = 'relu'))
model.add(Dense(num_classes, activation = 'softmax'))
# compile model
model.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])
model.fit(X_train, y_train, validation_data = (X_test, y_test), epochs = 10, batch_size = 200, verbose = 2)
scores = model.evaluate(X_test, y_test, verbose = 0)
print("MLP Error: %.2f%%" % (100-scores[1]*100))
Why one-hot?
Many machine learning algorithms cannot operate on label data directly. They require all input variables and output variables to be numeric. For e.g. red[1,0,0], green[0,1,0], blue[0,0,1].
CNN (Convolutional Nerual Network)
Mnist using CNN
import numpy
from keras.models import Sequential
from keras.layers import Dense, Dropout, Flatten
from keras.layers.convolutional import Conv2D
from keras.layers.convolutional import MaxPooling2D
from keras.utils import np_utils
from keras import backend as K
K.set_image_dim_ordering('th')
seed = 7
numpy.random.seed(seed)
data = numpy.load('mnist.npz')
X_train, y_train = data['x_train'],data['y_train']
X_test, y_test = data['x_test'], data['y_test']
X_train = X_train.reshape(X_train.shape[0], 1,28,28).astype('float32')
X_test = X_test.reshape(X_test.shape[0], 1,28,28).astype('float32')
# normalize inputs from 0-255 to 0-1
X_train = X_train/255
X_test = X_test/255
# convert to 'one-hot' format
y_train = np_utils.to_categorical(y_train)
y_test = np_utils.to_categorical(y_test)
num_classes = y_test.shape[1]
# create model
model = Sequential()
model.add(Conv2D(32,(5,5), input_shape = (1,28,28), activation = 'relu'))
model.add(MaxPooling2D(pool_size = (2,2)))
model.add(Dropout(0.2))
model.add(Flatten())
model.add(Dense(128, activation = 'relu'))
model.add(Dense(num_classes, activation = 'softmax'))
# compile model
model.compile(loss = 'categorical_crossentropy', optimizer = 'adam', metrics = ['accuracy'])
# fit the model
model.fit(X_train, y_train, validation_data = (X_test, y_test), epochs = 10, batch_size = 200, verbose = 2)
# evaluation of the model
scores = model.evaluate(X_test, y_test, verbose = 0)
print('CNN Error: %.2f%%' % (100 - scores[1]*100))
K.set_image_dim_ordering('th')
当图片维序类型为th时，输入数据格式为[samples][channels][rows][cols];
当图片维序类型为tf时，输入数据格式为[samples][rows][cols][channels];
channels代表图片的RGB通道，彩色图片为3，灰度图片为1
keras.layers.Conv2D(nums of filters, kernel_size, input_shape)
nums of filters actually it’s nums of neurons, since each neuron perform a different convolution on the input to the layer (more precisely, the neurons’ input weights form convolution kernels). A feature map is the result of applying a filter (thus, you have as many feature maps as filters), and its size is a result of window/kernel size of your filter and stride.
hahaha?
ddd
why
dfasd
