Hi

in C1,W3, I proofed myself all of 6 equations for one training example.

In the picture below, I am OK with all equations in left side. The thing I don’t get it is why in right side we sum Db?

besides , why we don’t sum dWs?

In the slide you presented, the right side equations are just the vectorized versions of the left side equations.

I believe you are talking about computing the right side (vectorized) of db[1]? This is just the vectorized version of the same equation for db[1] on the left side. Instead of the derivative for one sample in the left side, we sum up the derivatives for multiple samples on the right side.

Thanks for your response.