Reference no: EM133341704
Problem 1: Suppose we have two random variables, X and Y, and we draw a sample from each as below. Use a Wilcoxon rank sum test to test the null hypothesis that X and Y follow the same distribution. Please report your results with interpretation.
X : 26, 30, 20, 21, 15, 34, 22
Y : 27, 23, 33, 29, 28, 18, 40, 12
Problem 2: Below is a 2x2 contingency table for two categorical variables X, Y. Perform Fisher's exact test to determine whether or not X and Y are independent. Be sure to include everything necessary to do the test, including a set of hypotheses, a significance level of your choice, and a fully-explained conclusion.
|
Y = 0
|
Y = 1
|
X = 0
|
12
|
25
|
X = 1
|
3
|
7
|
Problem 3: Suppose that we have a classification problem with two classes, Y = 0 and Y = 1, and one feature measurement X . Suppose that
X|Y = 0 ∼ N(µ0, σ02)
X|Y = 1 ∼ N(µ1, σ12)
We will use 0-1 loss for this task. You may assume that a priori, both classes are equally likely.
1. Suppose that σ0 = σ1. Express the Bayes classifier in terms of the parameters. Express the Bayes error rate in terms of the parameters. You may leave your answer in the form of a sum of integrals for the Bayes error rate.
2. Suppose that µ0 = µ1, σ0 ≠ σ1. Express the Bayes classifier in terms of the param- eters. Express the Bayes error rate in terms of the parameters. You may leave your answer in the form of a sum of integrals for the Bayes error rate.
3. Suppose that µ0 = µ1, σ0 = σ1. Express the Bayes classifier in terms of the parameters. Express the Bayes error rate in terms of the parameters.