8 Tricks for Configuring Backpropagation to Train Better Neural Networks

版权声明：本文为博主原创文章，未经博主允许不得转载。 https://blog.csdn.net/hajungong007/article/details/87925253

• Stochastic Versus Batch Learning
• Shuffling the Examples
• Normalizing the Inputs
• The Sigmoid
  隐含层使用relu
• Choosing Target Values
  如果目标值是[0 , 1]，将值设为[0.1, 0.9]，避免目标值处于激活函数的临界值。
• Initializing the Weights
  Xavier weight initialization scheme.
• Choosing Learning Rates
  AdaGrad and Adam
• Radial Basis Function vs Sigmoid

ref：https://machinelearningmastery.com/best-advice-for-configuring-backpropagation-for-deep-learning-neural-networks/