Tensorflow2.1基礎知識---搭建神經網絡八股以及小案例實戰

本片文章的目的是:利用Tensorflow API tf.keras搭建網絡八股(六步法)

六步法：

1. 導入相關的模塊，也就是 import
2. 加載訓練集和測試集，也就是加載train（x_train數據、y_train標籤）、test（x_test數據、y_test標籤）數據
3. 前向傳播（搭建神經網絡結構，逐層描述每層網絡），也就是model = tf.keras.models.Sequential
4. 配置訓練時所用的方法（也就是優化器，損失函數，評測指標的選擇），也就是model.compile
5. 進行數據的訓練（告訴訓練集和測試集的輸入特徵和標籤，batch的值，以及迭代多少次數據集），也即是model.fit
6. 利用summary（）函數打印出網絡的結構和參數統計

對上述用到的tf.keras模塊中的函數進行進一步的介紹

1. model = tf.keras.models.Sequential([網絡結構]) #描述各層網絡
   網絡結構舉例：
   1. 拉直層：tf.keras.layers.Flatten()
   2. 全連接層：tf.keras.layers.Dense(神經元個數，activation=“激活函數”，kernel_regularizer=“正則化函數”)
      activation可選的字符串：“relu”、“softmax”、“sigmoid”、“tanh”
      kernel_regularizer可選：tf.keras.regularizers.l1()、tf.keras.regularizers.l2()
   3. 卷積層：tf.keras.layers.Conv2D（filters = 卷積核個數，kernel_size = 卷積核尺寸，strides = 卷積步長，padding = “valid”or“same”）
   4. LSTM層：tf.keras.layers.LSTM()
2. model.compile(optimizer=優化器，loss=損失函數，metrics=[“準確率”])
   1. Optimizer可選：
      i. ‘sgd’ or tf.keras.optimizers.SGD(lr=學習率,momentum=動量參數)
      ii. 'adagrad’or tf.keras.optimizers.Adagrad(lr=學習率)
      iii. 'adadelta’or tf.keras.optimizers.Adadelta(lr=學習率)
      iv. 'adam’or tf.keras.optimizers.Adam(lr=學習率,beta_1=0.9,beta_1=0.999)
   2. loss可選：
      i. ‘mse’ or tf.keras.losses.MeanSquaredError()
      ii. ‘sparse_categorical_crossentropy’ or tf.keras.losses.SparseCategoricalCrossentropy(from_logots=False)
   3. Metrics可選：
      i. ‘accuracy’:y_和y都是數值，如y_=[1] y=[1]
      ii. ‘categorical_accuracy’:y_和y都是獨熱碼（概率分佈）,如y_=[0,1,0] y=[0.256,0.695,0.048]
      iii. ‘sparse_categorical_accuracy’:y_是數值，y是獨熱碼（概率分佈）,y_=[1] y=[0.256,0.695,0.048]
3. model.fit(訓練集的輸入特徵，訓練集的標籤，batch_size=，epochs=，validation_data=(測試集的輸入特徵，測試集的標籤)，validation_split=從訓練集劃分多少比例給測試集，validation_freq=多少次epoch測試一次)
4. model.summary() 打印出網絡的結構和參數統計

案例實戰

• 案例1：利用tf.keras實現鳶尾花分類

  import tensorflow as tf
  from sklearn import datasets
  import numpy as np
  
  x_train = datasets.load_iris().data
  y_train = datasets.load_iris().target
  
  np.random.seed(116)
  np.random.shuffle(x_train)
  np.random.seed(116)
  np.random.shuffle(y_train)
  tf.random.set_seed(116)
  
  model = tf.keras.models.Sequential([
      tf.keras.layers.Dense(3,activation = 'softmax',kernel_regularizer = tf.keras.regularizers.l2())
  ])
  
  model.compile(optimizer = tf.keras.optimizers.SGD(lr = 0.1),
               loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),
               metrics = ['sparse_categorical_accuracy'])
  
  model.fit(x_train,y_train,batch_size = 32,epochs = 500,validation_split = 0.2,validation_freq = 20)
  
  model.summary()
  

  將上述代碼封裝成class

  import tensorflow as tf
  from tensorflow.keras.layers import Dense
  from tensorflow.keras import Model
  from sklearn import datasets
  import numpy as np
  
  x_train = datasets.load_iris().data
  y_train = datasets.load_iris().target
  
  np.random.seed(116)
  np.random.shuffle(x_train)
  np.random.seed(116)
  np.random.shuffle(y_train)
  tf.random.set_seed(116)
  
  class IrisModel(Model):
      def __init__(self):
          super(IrisModel,self).__init__()
          self.dl = Dense(3,activation = 'sigmoid',kernel_regularizer = tf.keras.regularizers.l2())
      def call(self,x):
          y = self.dl(x)
          return y
      
  model = IrisModel()
  
  model.compile(optimizer = tf.keras.optimizers.SGD(lr = 0.1),
               loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),
               metrics = ['sparse_categorical_accuracy'])
  
  model.fit(x_train,y_train,batch_size = 32,epochs = 500,validation_split = 0.2,validation_freq = 20)
  
  model.summary()
  

• 案例2：利用tf.keras實現mnist手寫數字識別

  import tensorflow as tf
  
  mnist = tf.keras.datasets.mnist
  (x_train,y_train),(x_test,y_test) = mnist.load_data()
  x_train,x_test = x_train / 255.0,x_test / 255.0
  
  model = tf.keras.models.Sequential([
      tf.keras.layers.Flatten(),
      tf.keras.layers.Dense(128,activation = 'relu'),
      tf.keras.layers.Dense(10,activation = 'softmax')
  ])
  
  model.compile(optimizer = 'adam',
               loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),
               metrics = ['sparse_categorical_accuracy'])
  
  model.fit(x_train,y_train,batch_size = 32,epochs = 5,validation_data = (x_test,y_test),validation_freq=1)
  model.summary()
  

  將上述代碼封裝成class

  import tensorflow as tf
  from tensorflow.keras.layers import Dense,Flatten
  from tensorflow.keras import Model
  
  mnist = tf.keras.datasets.mnist
  (x_train,y_train),(x_test,y_test) = mnist.load_data()
  x_train,x_test = x_train / 255.0,x_test / 255.0
  
  class MnistModel(Model):
      def __init__(self):
          super(MnistModel,self).__init__()
          self.flatten = Flatten()
          self.d1 = Dense(128,activation = 'relu')
          self.d2 = Dense(10,activation = 'softmax')
      def call(self,x):
          x = self.flatten(x)
          x = self.d1(x)
          y = self.d2(x)
          return y
      
  model = MnistModel()
  
  model.compile(optimizer = 'adam',
               loss = tf.keras.losses.SparseCategoricalCrossentropy(from_logits = False),
               metrics = ['sparse_categorical_accuracy'])
  
  model.fit(x_train,y_train,batch_size = 32,epochs = 5,validation_data = (x_test,y_test),validation_freq=1)
  model.summary()
  

下面的是筆者的微信公衆號，歡迎關注，會持續更新c++、python、tensorflow、機器學習、深度學習等系列文章