Deep Learning Specialization vs PyTorch for Deep Learning

Course Q&A Deep Learning Specialization Neural Networks and Deep Learning
, ,
1

Hi, what are the major differences between these two specializations?

Are they complementary courses? They cover similar topics, so I can skip one of the two, or would it be better to take them in a specific order?

Thanks,

Alfredo

2

I would recommend that you take DLS first: it really gives you the foundations for all the major types of ML models. It introduces all of them, explains what types of problems they can be used to solve and shows you how to actually create such implementations. In DLS they use TensorFlow as the solutions framework in courses C2, C4 and C5 (C3 is not a programming course). In each course Professor Andrew Ng will first show you how to create the fundamental algorithms in straight python and numpy and then you graduate to using TF for building full solutions. This is important from a pedagogical standpoint so that you get solid intuition about how the various models work, whereas in TF or PyTorch you can basically treat the core algorithms as “black boxes”.

The PyTorch Specialization is newer and it is structured very similarly to the TensorFlow Specializations (not surprising given that they were both created by Professor Laurence Moroney): it assumes you already know what the different types of ML models are and what they are good for and then shows you how to create implementations of them using PyTorch. So the emphasis is mostly on how to use PyTorch and it assumes you already have the base knowledge that you get from DLS.