The Student Room Group

Normalisation

Consider the multiple regression model y=Xβ+u\displaystyle y=X\beta +u where E(u)=0\displaystyle E(u)=0 and V(u)=σ2In\displaystyle V(u) = \sigma^2 I_n

where β\displaystyle \beta is a K×1\displaystyle K \times 1 vector, and X\displaystyle X is a full rank n×K\displaystyle n \times K non-stochastic matrix.

By using the fact that limni=1niα=limn1nxαdx\displaystyle \lim_{n\to\infty}\sum_{i=1}^n i^\alpha = \lim_{n\to\infty}\int_1^n x^\alpha \, dx for all α>0\displaystyle \alpha>0

Find a suitable normalisation for i\displaystyle \sum i and for i2\displaystyle \sum i^2 as n\displaystyle n \to \infty and calculate the limit of the normalised sums.


I'll be truthful here... I've absolutely no idea what's being asked let alone how to proceed. Any help?

Quick Reply

Latest