Deep Learning Quick Reference
上QQ阅读APP看书,第一时间看更新

How to plan a machine learning problem

When building a new neural network, I recommend following the same basic steps every time. 

Deep neural networks can get very complicated, very quickly. A little bit of planning and organization and greatly accelerate your workflow!

The following are the steps for building a deep neural network:

  1. Outline the problem you're trying to solve.
  2. Identify the inputs and outputs of the model.
  3. Choose a cost function and metrics.
  4. Create an initial network architecture.
  5. Train and tune the network.