Neural Network Architecture-

As a beginner in the field,

  • How can I pick the right architecture for my application?? Are there any guidelines?
  • After finishing the specialization, will I be able to understand and(maybe) implement some NN architectures that are found in research papers?

Thank you for the feedback

Hey @Lina_Hourieh,

Do check this thread out. The rule-of-thumb here is that there is no rule-of-thumb. The right architecture is often found using iterating over multiple possible architectures, and as you keep on gaining more and more experience, your intuition will often guide you as to how you should modify the existing architecture.

I think that’s a question that you can answer and not me. If you grasp the concepts completely and in-depth, then I don’t see any reason as to why you won’t be able to do this. Also, you will have Internet at your disposal, so whenever you won’t be able to figure out something, you can always do a quick search, and that will get you up to speed. And why to stop there only? Perhaps you could create a new neural network architecture that will become so popular one day that the DeepLearning.AI team will be forced to include that in the course itself, so, all the best for that :nerd_face:

Let me know if this helps.


1 Like