第二章介绍了三种情况下安装Keras的方法以及API,而且包括一些常用的一些操作。

安装

《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
当安装的Tensorflow是GPU版本的时候,Keras调用的时候会自动选择GPU版本的Tnesorflow作为后端进行运行

API

1、什么是Tensor?
多维矩阵
2、Keras定义模型的两种方法

  • Sequential
  • Functional

3、常用操作

  • Dense
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
  • Recurrent
  • simpleRNN
  • GRU
  • LSTM
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
  • Conv1D
  • Conv2D
  • MaxPooling1D
  • MaxPooling2D
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
  • Regularization
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
  • BatchNormalization
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
  • Activation
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
  • Loss function
    1、Error loss
    mse
    rmse
    mae
    mape
    msle
    2、Hingle loss
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
    3、Class loss
    binary cross-entropy
    categorical cross-entropy
  • Metrics
  • Optimizer
  • Saving and loading the weights and architecture
  • Callbacks
    1、Checkingpoint
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
    2、Using Tensorboard
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API
    3、Using Quiver
    《Deep Learning wih Keras》CHAPTER02 notes: Keras Installation and API

相关文章: