I still have problem understanding how to compute the “gradient”

So, on the left side, that is the derivative term of the cost function J, but how is that equal to f_wb - y?

Thank you so much!

I still have problem understanding how to compute the “gradient”

So, on the left side, that is the derivative term of the cost function J, but how is that equal to f_wb - y?

Thank you so much!

1 Like

The first term is the gradient for the bias ‘b’. It does not depend on ‘x’.

The second term is the gradient for the weight ‘w’.