why we write Matrix A in the shape 1 then transpose it to be in the shape 2, not write it the shape 2 from the beginning? ?
1 Like
Hi @hadeer.awad02, it’s because the convention is we use a row to represent a sample.
This is a matrix algebra example.
In machine learning, sometimes you need a matrix in one shape for one calculation, but need that same matrix in a different shape for a different calculation.
Transposition is an important skill.