Actuarial Outpost Search Results
 Register Blogs Wiki FAQ Calendar Search Today's Posts Mark Forums Read
 FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

 Showing results 1 to 25 of 300 Search took 0.14 seconds. Search: Posts Made By: daaaave
 07-09-2019, 12:29 PM Replies: 32 Views: 7,187 Posted By daaaave The rumor I've heard is 2020 is the goal date. The rumor I've heard is 2020 is the goal date.
 Forum: MAS-II 06-17-2019, 04:26 PM Replies: 16 Views: 4,671 Posted By daaaave Thanks for the quick clarification. Thanks for the quick clarification.
 Forum: MAS-II 06-17-2019, 03:16 PM Replies: 16 Views: 4,671 Posted By daaaave Final answer key is up. They changed #5 from D to... Final answer key is up. They changed #5 from D to E, #18 from A to E, and on #35 added E as an accepted answer choice.
 05-14-2019, 12:12 PM Replies: 4 Views: 1,300 Posted By daaaave You probably want to look under wherever they... You probably want to look under wherever they cover ILFs. The point is that in computing the pure premium for the 250,000 limit, we want to use only data coming from policies with a limit of at least...
 Forum: MAS-II 05-02-2019, 12:00 PM Replies: 16 Views: 4,671 Posted By daaaave I think by this you are thinking their expression... I think by this you are thinking their expression becomes \sigma_i = (\alpha + \beta x_i)^2 + 1 Is that what you are saying? The problem with that is that the link function has to be a...
 Forum: MAS-II 04-30-2019, 04:58 AM Replies: 16 Views: 4,671 Posted By daaaave I've updated my solution file above to correct a... I've updated my solution file above to correct a mistake I made in # 38.
 Forum: Non-Actuarial Topics 04-29-2019, 11:08 AM Replies: 12 Views: 505 Posted By daaaave No, a mixture of 2 normals isn't normal. For a... No, a mixture of 2 normals isn't normal. For a more extreme example, consider a 50/50 mixture of a normal with mean 1,000 and SD 1 with another normal with mean 2,000 and SD 1. The mixture is...
 Forum: MAS-II 04-29-2019, 10:51 AM Replies: 16 Views: 4,671 Posted By daaaave I'm attaching draft solutions in case anyone... I'm attaching draft solutions in case anyone wants them to help deciding what (if anything) they want to appeal. I disagree with the preliminary answer key on the following 3 problems: On #5, I...
 Forum: MAS-II 04-22-2019, 11:22 AM Replies: 16 Views: 4,671 Posted By daaaave The intent of the problem is to get at the idea... The intent of the problem is to get at the idea that model selection is really more of a frequentist topic, and is antithetical to the Bayesian approach. Picking a single model based on WAIC and...
 Forum: MAS-II 04-22-2019, 11:08 AM Replies: 55 Views: 21,886 Posted By daaaave I updated my solutions PDF from earlier in this... I updated my solutions PDF from earlier in this thread to clean up some typos and add more detail on some problems.
 04-22-2019, 11:03 AM Replies: 128 Views: 41,655 Posted By daaaave As you are calling this Example 3.1, you are... As you are calling this Example 3.1, you are using the 3rd edition of the Intro to Ratemaking text, not the 4th edition, which is the one on the syllabus. There are some topics in Chapter 5 that are...
 Forum: Non-Actuarial Topics 03-13-2019, 04:51 PM Replies: 35 Views: 850 Posted By daaaave TIL The Right doesn't properly appreciate steak. TIL The Right doesn't properly appreciate steak.
 01-08-2019, 02:55 PM Replies: 697 Views: 164,612 Posted By daaaave In the preferences of most R installations is an... In the preferences of most R installations is an option to select the default working directory. See e.g. https://support.rstudio.com/hc/en-us/articles/200549016-Customizing-RStudio for R-Studio. ...
 Forum: Probability 01-07-2019, 10:22 AM Replies: 6 Views: 1,144 Posted By daaaave Intuitively, independence means that knowing... Intuitively, independence means that knowing whether or not one event occurred doesn't affect your knowledge of whether or not the other one occurred. Here, P[A] = 1/2. If I told you that B occurred,...
 12-21-2018, 01:15 PM Replies: 1 Views: 849 Posted By daaaave Whenever X is a continuous loss variable, E[X... Whenever X is a continuous loss variable, E[X \wedge 20] = \int_0^{20} x \cdot f(x) \, dx + 20 \Pr[X>20] In their expression, they are finding Pr[X>20] by doing 1 - Pr[X<=20] as the integral for...
 Forum: MAS-II 11-06-2018, 08:00 PM Replies: 55 Views: 21,886 Posted By daaaave Thanks, I've updated my solution file to remove... Thanks, I've updated my solution file to remove numbers for those problems.
 Forum: MAS-II 11-06-2018, 03:46 PM Replies: 55 Views: 21,886 Posted By daaaave Okay, here are some solutions to the exam. I made... Okay, here are some solutions to the exam. I made some comments on #1 and #2, but can't compute numbers without the case study given on the exam.
 Forum: MAS-II 11-06-2018, 02:14 PM Replies: 55 Views: 21,886 Posted By daaaave I was assuming they used the same case study as... I was assuming they used the same case study as distributed before the sample questions, as both are about systolic blood pressure...
 Forum: MAS-II 11-06-2018, 11:03 AM Replies: 55 Views: 21,886 Posted By daaaave Some thoughts on the exam as the appeal deadline... Some thoughts on the exam as the appeal deadline is Friday: On #1, I get E from the numbers in the published case study, but my understanding is the case study given on the exam was different and...
 Forum: MAS-II 11-01-2018, 03:31 PM Replies: 55 Views: 21,886 Posted By daaaave Saying that we aren't yet stationary or mixing... Saying that we aren't yet stationary or mixing well is consistent with saying that we need a larger sample size. We aren't mixing well because the auto-correlation is strong enough that it takes a...
 Forum: MAS-II 11-01-2018, 09:03 AM Replies: 55 Views: 21,886 Posted By daaaave I agree with their answer. If you look at the... I agree with their answer. If you look at the examples in Figure 8.5 of McElreath, in the example with a flat prior that should be replaced by a weakly informative prior (which has a reduced...
 10-29-2018, 08:27 AM Replies: 2 Views: 818 Posted By daaaave Yes, it should. Thanks for pointing this out,... Yes, it should. Thanks for pointing this out, I'll update the lesson. The question is asking about annual claim amounts. One data point is thus one annual claim amount for 1 insured, and the...
 Forum: MAS-II 10-15-2018, 11:17 AM Replies: 55 Views: 21,886 Posted By daaaave For #12, I'm not sure what you mean by the... For #12, I'm not sure what you mean by the highest split. If you mean the fact that we are splitting first where X_1 and Y are highest, that is not the reason why the answer is B. The reason why that...
 Forum: MAS-II 10-15-2018, 09:53 AM Replies: 55 Views: 21,886 Posted By daaaave For #3, you used the fact that for Poisson-Gamma,... For #3, you used the fact that for Poisson-Gamma, the Buhlmann and Bayesian credibility estimates match and used the Bayesian approach, which is fine. I think it's also worth writing out how the...
 10-07-2018, 03:28 PM Replies: 3 Views: 582 Posted By daaaave In the example, our data point was 15,000,000 in... In the example, our data point was 15,000,000 in observed losses. That is our X-bar, and when we do Z * 15,000,000, we are doing Z * X-bar. The 10,000,000 of the prior estimate is our M, so the (1-Z)...
 Showing results 1 to 25 of 300

All times are GMT -4. The time now is 02:42 PM.

 -- Default Style - Fluid Width ---- Default Style - Fixed Width ---- Old Default Style ---- Easy on the eyes ---- Smooth Darkness ---- Chestnut ---- Apple-ish Style ---- If Apples were blue ---- If Apples were green ---- If Apples were purple ---- Halloween 2007 ---- B&W ---- Halloween ---- AO Christmas Theme ---- Turkey Day Theme ---- AO 2007 beta ---- 4th Of July Contact Us - Actuarial Outpost - Archive - Privacy Statement - Top