Let
X be a nXk matrix
Y,E and be nX1 vectors
X' is the transpose of X
^ is the estimate for
Now,
or Y= X+E
i. Using principle of ordinary least squares we minimise the sum of squared residuals :
E=Y- X (Residuals)
Sum of squared residuals: E'E=(Y- X^)' (Y- X^)= Y'Y-^'X'Y-Y'X^+^'X'X^ = Y'Y-2^'X'Y+^'X'X^
To minimise, we partially take the derivative with ^
Now,we get the normal equation as
(X'X)^=X'Y
or ^=((X'X)^-1)*X'Y ---1
ii.
Taking expectations on both sides of equation 1
E( ^)=E(((X'X)^-1)*X'Y)
=((X'X)^-1)X' E(Y)= ((X'X)^-1)X' (X+E)= ((X'X)^-1)X'X= (E(E)=0)
Hence E( ^)=
Therefore unbiasedness is proved
iii. Variance ^ = E(( ^-)( ^-)')= E( ((X'X)^-1)X'u)((X'X)^-1)X'u)')= E((X'X)^-1)X'uu'X(X'X)^-1)= (X'X)^-1)X'E(uu')X(X'X)^-1=(X'X)^-1)X'(sigma sq)X(X'X)^-1=(sigma sq)(X'X)^-1)X'X(X'X)^-1= (sigma sq)(X'X)^-1 which is the desired variance covariance matrix.
II. Derivations (You must show all your work for full credit.) i. Given the model y=XB+ɛ,...
4. Consider the linear model Y = XB+e, where e MV N(0,021). (1) Derive the formula for , the least square estimate of B, using the matrix notation (2) Show that ß is an unbiased estimate for B. (3) Derive the formula for var(), using matrix notation.
Consider the following linear regression model 1. For any X = x, let Y = xB+U, where B erk. 2. X is exogenous. 3. The probability model is {f(u; ) is a distribution on R: Ef [U] = 0, VAR; [U] = 62,0 >0}. 4. Sampling model: {Y}}}=1 is an independent sample, sequentially generated using Y; = xiß +Ui, where the U; are IID(0,62). (i) Let K > 0 be a given number. We wish to estimate B using least-squares...
Problem 3: Absence of Intercept Consider the regression model Y, = BX,+", where , and X, satisfy Assumptions SLR1-SLR5. Y (i) Let B denote an estimator of B that is constructed as P where Y and X as are the sample means of Y,and X,, respectively. Show that B is conditionally unbiased. Derive the least squares estimator of B. Show that the estimator is conditionally unbiased. Derive the conditional variance of the estimator. (ii) (iii) (iv) 2
2. The linear regression model in matrix format is Y Χβ + e, with the usual definitions Let E(elX) 0 and T1 0 0 01 0 r2 00 0 0 0 0.0 0 γΝ 0 00 Notice that as a covariance matrix, Σ is bymmetric and nonnegative definite () Derive Var (0LS|x). (ii) Let B- CY be any other linear unbiased estimator where C' is an N x K function of X. Prove Var (BIX) 2 (X-x)-1 3. An oracle...
Suppose we have the full rank linear model y = XA+ Ewiun xp design matrix X, normal errors E N (0,0?Inxn). Let b be the least squares estimator of B. (C) Prove that (b-B)? XT X(6-8) o2 follows the x? distribution. Hint: Write Xb in terms of X, B and e. (d) Hence derive a 100(1 - a)% joint confidence region of ß given in notes (b - B) TXTX(b-)/po<Fa:pon-p, where Faip,n-p denotes the upper ath quantile of the Fpin-p...
please help Question 2. (2.5 points. You are considering the model Y = XB + X2B, +€, where E(€) = 0 and E(ee') = oʻI,.. Here, X, is n xp and X, is n xq, where p >1 and q> 1. Suppose that in fact, unknown to you, B, = 0. In other words, (*) is an over-parameterized model. Let e be the vector of residuals corresponding to the fitted version of *) based on the least squares method. Does...
Q4.. [40 points] Consider the multiple linear regression model given by y - XB -+ s, where y and e are vectors of size 8 × 1, X ls a matrix of size 8 x 3 and Disa vector of sze 3 × 1. Also, the following information are available e = 22 y -2 and XTy 3 1. [10 points) Estimate the regression coefficients in the model given above? 2. [4 points] Estimate the variance of the error term...
Q. 1 Consider the multiple linear regression model Y = x3 + €, where e indep MV N(0,0²V) and V +In is a diagonal matrix. a) Derive the weighted least squares estimator for B, i.e., Owls. b) Show Bwis is an unbiased estimator for B. c) Derive the variances of w ls and the OLS estimator of 8. Is the OLS estimator of still the BLUE? In one sentence, explain why or why not.
2. Given the following model: Y, = B. +X;B, + Mi a. Suppose we estimate the model ignoring the constant term. Show that the resulting estimator (call it ß, ) is biased. b. Derive the variance
3. Let y = (yi..... Yn) be a set of re- sponses, and consider the linear model y= +E, where u = (1, ..., and e is a vector of zero mean, uncorrelated errors with variance o'. This is a linear model in which the responses have a constant but unknown mean . We will call this model the location model. (a) If we write the location model in the usual form of the linear model y = X 8+...