01.08.2014 Views

Estimation, Evaluation, and Selection of Actuarial Models

Estimation, Evaluation, and Selection of Actuarial Models

Estimation, Evaluation, and Selection of Actuarial Models

SHOW MORE
SHOW LESS

Create successful ePaper yourself

Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.

100 APPENDIX A. SOLUTIONS TO EXERCISES<br />

Exercise 8 Ĥ(12) = 2 15 + 1<br />

12 + 1 10 + 2 6<br />

e −0.65 =0.522.<br />

=0.65. The estimate <strong>of</strong> the survival function is Ŝ(12) =<br />

Exercise 9 The information may be organized as in the following table:<br />

age(t) #ds #xs #us r Ŝ(t)<br />

0 300<br />

294<br />

1 6 300<br />

300 =0.98<br />

2 20<br />

3 10 314 0.98 304<br />

314 =0.94879<br />

4 30 10 304 0.94879 294<br />

304 =0.91758<br />

5 a 324 0.91758 324−a<br />

324<br />

=0.892 =⇒ a =9<br />

7 45<br />

9 b 279−a = 270 0.892 270−b<br />

270<br />

10 35<br />

12 6 244−a − b =235− b 0.892 270−b 229−b<br />

270 235−b<br />

=0.856 =⇒ b =4<br />

13 15<br />

Exercise 10 Ĥ(t 10) − Ĥ(t 9)=<br />

n−9 1 Ĥ(t 3 )=<br />

22 1 + 21 1 + 20 1 3 )=<br />

e −0.14307 =0.8667.<br />

Exercise 11 0.60 = 0.72 r 4−2<br />

r 4<br />

, r 4 =12. 0.50 = 0.60 r 5−1<br />

r 5<br />

, r 5 =6. With two deaths at the fourth<br />

death time <strong>and</strong> the risk set decreasing by 6, there must have been four censored observations.<br />

Exercise 12 In order for the mean to be equal to y, wemusthaveθ/(α − 1) = y. Letting α be<br />

arbitrary (<strong>and</strong> greater than 1), use a Pareto distribution with θ = y(α − 1). This makes the kernel<br />

function<br />

α[(α − 1)y]α<br />

k y (x) =<br />

[(α − 1)y + x] α+1 .<br />

Exercise 13 The data points <strong>and</strong> probabilities can be taken from Exercise 4. They are:<br />

y j p(y j )<br />

0.1 0.0333<br />

0.5 0.0323<br />

0.8 0.0311<br />

1.8 0.0623<br />

2.1 0.0300<br />

2.5 0.0290<br />

2.8 0.0290<br />

3.9 0.0557<br />

4.0 0.0269<br />

4.1 0.0291<br />

4.8 0.0916<br />

The probability at 5.0 is discrete <strong>and</strong> so should not be spread out by the kernel density estimator.<br />

Because <strong>of</strong> the value at 0.1, the largest available b<strong>and</strong>width is 0.1. Using this b<strong>and</strong>width <strong>and</strong> the<br />

triangular kernel produces the following graph.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!