00:01So here using this graph of f shown in this figure, we need to evaluate the integral by interpreting its geometry.
00:09So here evaluate.
00:13So evaluation of line...
Question
This problem uses MATLAB. Please do not attempt this problem if you do not have the program, and please include your code and the figure you find through the code. I have included my code to help. Thank you for your time.
Included MATLAB Code
%PROBLEM 4
%TRUE MODEL PARAMETERS
mu = [50 75]; %TRUE means
C = [98 22; 22 15]; %TRUE covariance matrix
nsim = 10^5; %Number of simulations
mhat = zeros(1, nsim); %Initialize the mhat data array
%==============================================
for j = 1:nsim
n = 100; %Sample size
xy = mvnrnd(mu, C, n); %Compute an n x 2 array of (speed, noise) data
x = xy(:, 1);
y = xy(:, 2);
% ESTIMATED MODEL PARAMETERS:
% INSERT YOUR CODE HERE:
PROBLEM 4 (25 pts) Suppose that you present your prediction model to management as, say, the model?
Now, you could answer by giving numbers related to the residual sum of squares or the r2 value. However, most managers would have no idea what you're talking about, and hence, would likely be turned off. But even if they did know what you're talking about, these numbers do not answer the question. To clarify the question, management might say, "OK. Your model tells me that for each 1 mph increase in the vehicle speed, you predict an increase of 0.22 dBA in the cabin noise. Hence, in going from 50 mph to 80 mph, the cabin noise will increase by ~7 dBA. Since it is in this speed range that the noise becomes more irritating, it would seem that we could definitely gain market value by developing a sound system controller. After all, 7 dBA is close to a doubling of the perceived noise level. But, what IF that slope is actually 0.1? If that's true, then there would only be an increase of 3.5 dBA; an increase that most people wouldn't be bothered too much by it. We would have spent millions of dollars and not gained any real market value.
Actually, that manager is spot on! So, the question is, to put it in the context of this course: What is the at best (even for a Ph.D. student in statistics). Hence, in this problem, you will answer it by using simulations. The MATLAB code in the Appendix To run nsim = 10 simulations, I have started the code in the for PROBLEM 4.
(a) (5pts) Use the truth model information to compute the true values of m and b Solution:
(b) (20pts) (i) Complete the code to arrive at a histogram-based pdf for the estimator m. Then (ii) give the is an unbiased estimator for m. Finally, (iv) use the normal model to find the values such that Pr[-<ms+] = 0.9 Solution:
Figure 4(b) Simulation-based and normal model pdfs for m
Read More...
Close
Want better grades, but can’t afford to pay for Numerade?
Hmmm, doesn't seem like you have any playlists. Please add your first
playlist.
Create a New Playlist
`
What our students say
Adama Hogking
"Numerade is far more than just what we see. I mean it provides detailed explanation to a question with a written solution which is coupled with an audio clarification. I love this app. All I can do now is to give it a five stars."
Yahtzee7k
"In school? You need Numerade. It has made my life SO much easier. I can’t say enough good things about this app. Love it, in a literal sense."
mohammed ihsaan
"Really a wonderful, amazing and very useful app for the all types of students like engineering, science,etc .It provide more accurate answer and it give all the answers even no one websites can do. Really a useful thing to keep everyone."