STA462562 R HW 2 20pts Name
In this homework, you will apply knowledge from Chapters 910 in the analysis of the real data. You will also learn how to develop a power curve for a given hypothesis test. A failure to submit both RMD and HTML or pdf files on Canvas will automatically result is a 50 penalty. Project grading will be based on a rubric provided in Canvas.
Q1. 10pts
The following data set contains the insurance severity clams for Norwegian Fire losses. Use the following command in R to read the data. You need to first download the data from Canvas.
xscannorwegiancar.txt
The data set was used in the paper titled Focused selection of the claims severity distribution by Yinzhi Wang and Hobaek Haff, published in Scandinavian Actuarial Journal in 2019. This paper is posted on Canvas for your convenience. Fit the following distributions in R, using the method of Maximum Likelihood Estimation. For each distribution find MLEs and the value of the maximum loglikelihood function; then, plot the density function on the top of the data histogram.
a Exponential
b Gamma ,
c Lognormal ,
d Weibull ,
e Pareto ,
Note: Use the parametrization of these distributions as defined in R. Recall that we discussed this parametrization in class for the Gamma distribution.
Q2. 7pts Suppose you want to test the following hypotheses: Ho: 30,000
Ha: 30,000
Use 0.05 and 0.01.
a Define a rejection region for a given level. Assume 5,000. Note: the sample size is small!
b Construct two Power Curves for various values ofType II error rate. Note: you will need to consider various values for the sample mean.
c Reflect on your findings. How does level impact the shape of the power curve? How does power changes with ? How does power changes with ? A minimum 3 full sentences are required for full credit.
1
Q3. 3pts Reflect on the overall HW using minimum 5 full sentences. Discuss how this HW helped you understand better concepts from Ch 9 and Ch 10? Why is this HW different from the general inclass problem solving activities? Have you experienced any challenges in completing this HW?
2
Reviews
There are no reviews yet.