

FlashChat  Actuarial Discussion  Preliminary Exams  CAS/SOA Exams  Cyberchat  Around the World  Suggestions 


Thread Tools  Search this Thread  Display Modes 
#52




Quote:
__________________
FCAS 
#53




Just read through GLMs w/ Examples in R as a potential textbook to use for a statistics minor at my old school. I really like the Springer series books (Intro Statistical Learning, Applied Predictive Modeling, and Elements of Statistical Modeling).
I'd definitely suggest it if you want to get more in depth into the mathematics behind GLMs. It covers the Tweedie distribution and the R code is still up to date (Applied Predictive Modeling is pretty outdated for the version I have). At 500 pages, I'll probably refer back to this instead of the CAS monographs. My gf is taking Exam 8 and thinks it's easier to follow than the monograph. Even when it gets to the Linear Algebra, it's not that hard to parse though what's going on. Compared to Elements of Statistical Learning which is very dense. Last edited by Actuarially Me; 08122019 at 12:55 PM.. 
#54




Thanks for writing this up. I love working in R and all things tidy, so naturally I'm frustrated that there doesn't seem to be a ton of options when it comes to elastic net regression in R.
glmnet is limited on the distributions you can select (e.g. no gamma), and HDtweedie doesn't seem to allow you to specify the link (instead enforcing the canonical link for a given $p$). So how does one build an elastic net Gamma regression in R with a log link? 
#55




Quote:
__________________
FCAS 
Thread Tools  Search this Thread 
Display Modes  

