SOA/ASA/PA 기출 및 내용정리 - 23.10시험 Task3~4(기록용)

Task3

(a) Explain three differences between fitting a normal linear regression to log(X) compared to fitting a GLM with a log link function to the unaltered X variable.

 - prior : lognormal model, latter : GLM with a log link

 (1) The normal linear regression has a log tranformation applied to the response variable, and the GLM does not. The log transformation is reasonable for a variable that has right-skew.

 (2) The GLM has flexibility to select a probability distribution that best fits the shape of the response variable, whereas the normal linear regression model only allows for one distribution.

 (3) In the normal linear model the variance of the (transformed) response variable is constant while in the GLM the variance can be a function of the mean. > 익숙치 않다

 

(b) residual plot을 분석하라는데... 분석은 됬고 관련개념.

Homoscedasticity(등분산성) and heteroscedasticity(이분산성)

Wikipedia

Task4

(a) Absolute error, bagging example

	First Tree	Sencond Tree	Bagging Ensemble
Prediction	43000	40000	Avg(43000,40000)=41500
Actual	32084	32084	32084
Absolute Error	10916	7916	41500-3084=9416

 

https://www.youtube.com/watch?v=tjy0yL1rRRU

DataMListic

Bagging/Boosting 참조용...