In the course for the sigmoid function 1/1+e^-z we substitute z as z = mx+c
But on various sites instead of substituting z direct input is fed where z=x.
Why so?
Hi @Arya_Gaikwad ,
In the context of this course, z is the output of the linear transformation, as seen by the formula z = mx+c
By applying sigmoid function to z, it would squash the values in z between 0 and 1.
If sigmoid is applied to the input x, then it must be for other purposes.
1 Like
Sometimes the bias value is included in the weight matrix, by adding a constant feature value of 1 and adding an element to the weight vector.
Then you’d have only have z = X * w as the equation.