代做STAT7005 Multivariate Methods (Fall 2024) Assignment 1代做Statistics统计
- 首页 >> Algorithm 算法DEPARTMENT OF STATISTICS AND ACTUARIAL SCIENCE
STAT7005 Multivariate Methods (Fall 2024)
Assignment 1
Due: October 2, 2024 via Moodle
1. The file cork.txt has a 28 × 4 data matrix on the weights of bark deposits of 28 trees in the four directions, in the order of north (N), east (E), south (S), and west (W).
(a) For each of the four variables, check for univariate normality.
(b) Check for the multivariate normality of all variables by means of a Q-Q plot on squared Mahalanobis distances.
(c) Assuming multivariate normality, test the nullhypothesisthatthe mean weights of bark deposits are the same in all directions at the 5% significance level.
(d) Test, at the 5% significance level, the joint hypothesis that the mean weights of bark deposits in the N and S directions are equal, and those in the E and W directions are equal.
(e) Construct a 95% confidence interval for the contrast of means N + E − S − W, and draw a relevant conclusion about your findings.
2. The blackbody CIE (Commission Internationale de l’Éclairage) chromaticity spec- ification for a color temperature of 4000K is μ1 = 0.3804 and μ2 = 0.3768. In 10 color-matching trials, one subject had mean chromaticity values x(¯)1 = 0.3745 and x(¯)2 = 0.3719, and sample covariance matrix
Assume the sample comes from a bivariate normal population. Test, at the 1% sig- nificance level, the null hypothesis that the observations satisfy the 4000K standards, i.e. μ1 = 0.3804 and μ2 = 0.3768.
3. Two brands of machines are used to produce four types of products. For each brand, 10 machines were chosen at random and the times needed (in minutes) to produce each product are recorded in Table 1 and saved in times.txt.
(a) In R, create a data frame. in a suitable form. useful for analysis.
(b) Let μA and μB be the mean vectors for the times needed to produce the four products for each brand.
i. Assuming that the covariance matrices for the two brands are equal, test the hypothesis H0 : μA = μB at the 5% significance level.
ii. Repeat part (b)(i) but without assuming the covariance matrices are equal.
(c) A company wants to buy some machines for each brand and would therefore like to know the average performance for the two brands. In addition, Product 1 and Product 2 belong to the same product line (say, Line I),while Product 3 and Product 4 belong to another product line (say, Line II). Rather than considering the performance for each product separately, the company wants to know the performance for each product line instead. In particular, the company wants to test the hypothesis of the form
H0 : C(μA + μB) = (60, 50)>, (1)
where 60 minutes is the time being put to the test for Line I, and 50 minutes for Line II. Assume equal covariance matrices in this part.
i. Write down the appropriate C matrix for conducting this test.
ii. At the 5% significance level, test the hypothesis (1).
iii. Construct 90% Bonferroni confidence intervals for the mean times needed for Line I and Line II.
Brand |
Machine |
Time to produce product (minutes) |
|||
Product 1 |
Product 2 |
Product 3 |
Product 4 |
||
A |
1 |
19 |
45 |
22 |
27 |
2 |
20 |
38 |
25 |
20 |
|
3 |
17 |
42 |
32 |
16 |
|
4 |
14 |
29 |
27 |
15 |
|
5 |
18 |
36 |
28 |
18 |
|
6 |
18 |
38 |
24 |
20 |
|
7 |
23 |
50 |
30 |
26 |
|
8 |
24 |
55 |
40 |
29 |
|
9 |
16 |
39 |
42 |
22 |
|
10 |
17 |
34 |
38 |
20 |
|
B |
1 |
13 |
27 |
33 |
24 |
2 |
14 |
33 |
35 |
25 |
|
3 |
18 |
42 |
32 |
24 |
|
4 |
13 |
36 |
31 |
18 |
|
5 |
27 |
35 |
29 |
29 |
|
6 |
22 |
30 |
26 |
27 |
|
7 |
16 |
44 |
33 |
25 |
|
8 |
19 |
42 |
43 |
26 |
|
9 |
15 |
38 |
37 |
16 |
|
10 |
16 |
45 |
32 |
19 |
Table 1: Times needed to produce four types of products