Estimation, Evaluation, and Selection of Actuarial Models
Estimation, Evaluation, and Selection of Actuarial Models
Estimation, Evaluation, and Selection of Actuarial Models
You also want an ePaper? Increase the reach of your titles
YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.
119<br />
With regard to assumption (ii) <strong>of</strong> Theorem 3.29,<br />
Z θ<br />
0<br />
Z<br />
∂ 1 θ<br />
∂θ θ dx = −θ −2 dx = − 1<br />
0<br />
θ 6=0.<br />
Exercise 69 From Exercise 67 we have ˆα =0.55616, ˆθ =2, 561.1 <strong>and</strong> covariance matrix<br />
· ¸<br />
0.021503 −99.0188<br />
.<br />
−99.0188 1, 045, 668<br />
The function to be estimated is g(α, θ) =αθ with partial derivatives <strong>of</strong> θ <strong>and</strong> α. The approximated<br />
variance is<br />
£ ¤ · ¸· ¸<br />
0.021503 −99.0188 2, 561.1<br />
2, 561.1 0.55616 = 182, 402.<br />
−99.0188 1, 045, 668 0.55616<br />
The confidence interval is 1, 424.4 ± 1.96 √ 182, 402 or 1, 424.4 ± 837.1.<br />
Exercise 70 The partial derivatives <strong>of</strong> the mean are<br />
∂e µ+σ2 /2<br />
∂µ<br />
∂e µ+σ2 /2<br />
= e µ+σ2 /2 = 123.017<br />
= σe µ+σ2 /2 =134.458.<br />
∂σ<br />
The estimated variance is then<br />
£ ¤ · ¸· ¸<br />
0.1195 0 123.017<br />
123.017 134.458 = 2887.73.<br />
0 0.0597 134.458<br />
Exercise 71 The first partial derivatives are<br />
∂l(α, β)<br />
∂α<br />
∂l(α, β)<br />
∂β<br />
= −5α − 3β +50<br />
= −3α − 2β +2.<br />
The second partial derivatives are<br />
∂ 2 l(α, β)<br />
∂α 2 = −5<br />
∂ 2 l(α, β)<br />
∂β 2 = −2<br />
∂ 2 l(α, β)<br />
∂α∂β<br />
= −3<br />
<strong>and</strong> so the information matrix is · 5 3<br />
3 2<br />
¸<br />
.