C1_W3_Lab02_Sigmoid_function_Soln - How is 'z' calculated?

Course Q&A Machine Learning Specialization Supervised ML: Regression and Classification
1

β€œNote, the orange line is β€˜π‘§β€™ or 𝐰⋅𝐱(𝑖) + 𝑏 above. It does not match the line in a linear regression model.”

I don’t get how β€˜z’ is calculated and why it’s different from a linear regression model. Where do the β€˜w’ and β€˜b’ values come from to calculate β€˜z’ in this example?

Are we using linear regression to find the values β€˜w’ and β€˜b’, but each iteration of the algorithm performs the sigmoid function on β€˜z’?

2

We’re using gradient descent to find the best w and b values.

The gradients for logistic regression look quite similar to those for linear regression, other than including sigmoid() when computing f_wb.

3

Is the sigmoid function what makes z turn out different in this lab?

4

Yes.
Via some calculus magic, the form of the gradients are the same as for the linear regression cost function, if viewed in terms of the f_wb value.