Generate your own data to simulate the linear regression model y = 07x +e. Generate n such independent training data vectors. Also generate ntest independent data vectors for the testing step. Let N(u, 2) denote the Gaussian distribution with mean u and covariance E. Do this as follows: (i) Fix 0 once as specified below. (ii) Training data generate: For i =1 to n, generate x(i) ~ N(0,1), eli) ~ N(0,02I) and y(i) = 0 + xli) +eli), (iii) Test data generate: In a different loop, repeat (b) for i = 1 to Ntest. Use this data in the testing step and not for training. Question 1 (30 points]: Generate the data for the following settings: (a) [5 points] Pick n = 30,100,1000, d=5, and set e = 0. Let @ = [0,1,4, 2, 10, 23). (b) [5 points] Pick n = 30,100,1000, d = 5, and set 02 = 10-6. Let 6 = [0,1,4,2, 10, 23). (c) [5 points] Pick n = 100,100,1000, d =5, and set 04 = 10-6. Let 0 = [0,1,4,2, 10,23). (d) [5 points] Repeat (b) with generating x(i) ~ N(Mx,I) with x = [1,1,1,1,1]. For this you need to choose a non-zero 00. (e) [10 points] Repeat (b) with generating x(i) ~ N(0, 2) with being a general covariance matrix. Specify in your report what diagonal entries you chose for and what effect that had. Pick anything other than all equal entries