How parameter w is 4x3 matrix?

Hi @khirman,

Each of the vectors w_{i} is a 3x1 vector. So w_{i}^{T} is 1x3.

You are stacking 4 of those along the first axis. Thus you get a 4x3 matrix!

Hope this helps!

Hi @khirman , w is a 4x3 matrix because you have 3 input parameters (x1, x2, x3) and 4 logistic regression units (a1[1], a2[1], a3[1], a4[1]). Does this help in explaining?