In closed form solution of linear regression, in the formula of theta, if we open the inverse bracket(applying the inverse property), we will get the final answer as X-1.Y . Also we will get the same solution in closed form solution of locally weighted regression. So, while building the model, can I use this formula which I have suggested or should stick with previous one?
Doubt in Closed Form Solution
hey @abhaygarg2001,
Yes you are right about opening the inverse brackets using the inverse property.
But about using the updated formulae, for that you need to give it a try . It looks like it will work.
But i guess it becomes the same as of simple linear regression. The weights will loose there importance.
Just give it a try and see what happens.