Question: Best linear prediction. Suppose X; Y 2 L2 and (without loss of generality) V.X/ D 1. What is the best approximation of Y by an
Best linear prediction. Suppose X; Y 2 L2 and (without loss of generality) V.X/ D 1.
What is the best approximation of Y by an affine function a C bX of X? Show that the quadratic deviation E
.Y abX/2
is minimised by b D Cov.X; Y / and a D E.Y bX/. What does this mean when X; Y are uncorrelated?
Step by Step Solution
There are 3 Steps involved in it
Get step-by-step solutions from verified subject matter experts
