Question

linear stat modeling & regression

please ,

i need the solution for Q3, but i copy Q2 because you need info from Q2 in order to answer Q3.

2) Suppose you have multiple regression set up YxXBp The ridge regression estimator is given by Here, llell'-Σ.< where is a vector of Vik. a) Find the expectation and variance-covariance matrix of Bridge, when X'X is a diagonal matrix with each diagonal entry is eqal to. Com pare these variances with the variances of the estimated coefficients for the ordinary least square (OLS) estimator. b) Is ridge regression estimator unbiased for true parameter B if the pro- posed model is correct? (An estimator θ of θ is unbiased if E(0)-. ) c) An estimator θ, of θ based on n observations is called consistent mean squared sense if E(0 0)0 as n goes to infinity. For any estimator θ of θ, E(0-0). (Bias(0))2 + Var(0). proved but you can assume it to be true (Can be Let the ridge regression estimate ridg based on n observations be denoted band Bri Show Brp nis consistent in mean squared sense (that means , th component of Bridge is consistent estimator of β). 3) Find the Maximum Likelihood Estimator of β and σ2 for question 2. Hence, find the expression for AIC and BIC.

Answer 1

Cov ( と = 오K es cenne ," er@si)svar (A) CamScanner

B A Cs Scanned with CamScanner