Stat Assignment - Help

(Questions)

An insurance analyst is interested in predicting automobile insurance loss amounts. The variable descriptions are given

below:

loss This is the insurance loss amount recorded in dollars. This is the response variable.

X1, X2, X3 These are the three numeric predictors. You may assume they are unit-less. The data comes from

https://www.kaggle.com/c/allstate-claims-severity. Allstate chose not to reveal the nature of these

variables. All we know is that these predictors may be useful in predicting losses.

The population model of interest is:

where£ represent the iid normally distributed error terms with a standard deviation of oE, The data are split into train

and test data, with train getting exactly 60% of the data. Answer the following questions using the outputs given below.

The outputs and the table below used only the train data. The table given below the outputs gives some summary

measures for all the variables in the train data. For example, the mean and standard deviation of X1 in the train data are

0.50 and 0.20 respectively. You may assume all the multiple regression assumptions are met. Some output values have

been rounded or covered on purpose. Questions 1-6 are all about the train data. In questions 7 and 8 you consider the

test data.

i: Parameter Estimates Term Estimate Std Error t Ratio Prob>ltl

Intercept 775 352 2.20 0.0277* x1 2324 610 3.81 0.0001*

x2 2130 693 3.07 0.0021*

x3 1063 284 3.74 0.0002*

x1*x2 -1903 1203 -1.58 0.1138

Minimum

Mean

Standard Deviation

Maximum

(1) What is the estimated value of oE?

A) 2884

8) Cannot be determined.

C) 2344

D) 2324

E) 2760

I Summary of Fit

RSquare

RSquare Adj

Root Mean Square Error

Mean of Response

Observations (or Sum Wgts)

Loss

138 0 0 0

0.50 0.50 0.50

0.20 0.20 0.20

1.00 1.00 1.00

(2) What is the test statistic value for Ho: 84 = -1903 versus Ha: 84 '# -1903?

A) -1.58

8) 1.58

C) 0.1138

D) -3.16

E) 0.00

0.03

0.03

2784.11

3044.84

3000.00

Which model should the analyst choose? A) Model 1 B) Model 2 C) Model 3 D) Model 4 E) Model 5