in the previous video h(xi) was equal to theta knot + theta X which is also equal to equation of line.( at 2.12 whole square was introduced)
How h subscript theta (xi) = theta knot + theta X whole square?
Hello Sahibjot,
There was a slight mistake at the 2:12 timestep. But
it wasn’t considered while computing the gradients, thus doesn’t make much of a difference in the end. Though, the loss function should not have that square term over h(xi). Please neglect it.
Thank you for pointing it out.
I hope I was able to clear your doubt?