1.5 神经元实现

• 分拆数据集

  def load_data(filename):
      """read data from data file."""
      with open(filename, 'rb') as f:
          data = pickle.load(f, encoding='bytes')        return data[b'data'], data[b'labels']# trensorflow.DataSetclass CifarData:
      def __init__(self, filenames, need_shuffle):
          all_data = []
          all_labels = []        for filename in filenames:
              data,labels = load_data(filename)            for item,label in zip(data,labels):                if label in [0,1]:
                      all_data.append(item)
                      all_labels.append(label)
          self._data = np.vstack(all_data)        # 归一化，将0-255的数归一成0-1直接的数
          self._data = self._data / 127.5 - 1 
          self._labels = np.hstack(all_labels)
          self._num_examples = self._data.shape[0]
          self._need_shuffle = need_shuffle
          self._indicator = 0
          if self._need_shuffle:
              self._shuffle_data()        
      def _shuffle_data(self):
          # 混排 [0,1,2,3,4,5] -> [2,1,4,0,3,5]
          p = np.random.permutation(self._num_examples)
          self._data = self._data[p]
          self._labels = self._labels[p]    
      def next_batch(self, batch_size):
          """return batch_size examples as a batch."""
          end_indicator = self._indicator + batch_size        if end_indicator > self._num_examples:            if self._need_shuffle:
                  self._shuffle_data()
                  self._indicator = 0
                  end_indicator = batch_size            else:                raise Exception("have no more examples")        if end_indicator > self._num_examples:            raise Exception("batch size is lager then all examples")
          batch_data = self._data[self._indicator:end_indicator]
          batch_labels = self._labels[self._indicator:end_indicator]
          self._indicator = end_indicator        return batch_data, batch_labels
          
  train_filename = [os.path.join(CIFAR_DIR,'data_batch_%d' % i) for i in range(1,6)]
  test_filenames = [os.path.join(CIFAR_DIR, 'test_batch')]
  
  train_data = CifarData(train_filename, True)
  test_data = CifarData(test_filenames, False)
  
  batch_data,batch_labels = train_data.next_batch(10)

• 测试算法准确率

  init = tf.global_variables_initializer()
  batch_size = 20train_steps = 100000test_steps = 100with tf.Session() as sess:
      sess.run(init)    for i in range(train_steps):
          batch_data, batch_labels = train_data.next_batch(batch_size)
          loss_val, acc_val, _ = sess.run(
              [loss, accuracy, train_op],
              feed_dict={
                  x: batch_data,
                  y: batch_labels})        if (i+1) % 500 == 0:            print ('[Train] Step: %d, loss: %4.5f, acc: %4.5f' \
                  % (i+1, loss_val, acc_val))                
          if (i+1) % 5000 == 0:
              test_data = CifarData(test_filenames, False)
              all_test_acc_val = []            for j in range(test_steps):
                  test_batch_data, test_batch_labels \
                      = test_data.next_batch(batch_size)
                  test_acc_val = sess.run(
                      [accuracy],
                      feed_dict = {
                          x: test_batch_data, 
                          y: test_batch_labels
                      })
                  all_test_acc_val.append(test_acc_val)
              test_acc = np.mean(all_test_acc_val)
              print('[Test ] Step: %d, acc: %4.5f' % (i+1, test_acc))

作者：Meet相识_bfa5
链接：https://www.jianshu.com/p/fed88fcd3428