Actuarial Outpost
Go Back   Actuarial Outpost > Actuarial Discussion Forum > General Actuarial
FlashChat Actuarial Discussion Preliminary Exams CAS/SOA Exams Cyberchat Around the World Suggestions

General Actuarial Non-Specific Actuarial Topics - Before posting a thread, please browse over our other sections to see if there is a better fit, such as Careers - Employment, Actuarial Science Universities Forum or any of our other 100+ forums.

Thread Tools Search this Thread Display Modes
Old 04-14-2019, 10:32 AM
Colymbosathon ecplecticos's Avatar
Colymbosathon ecplecticos Colymbosathon ecplecticos is offline
Join Date: Dec 2003
Posts: 6,098

Originally Posted by magillaG View Post
But the recent "AI" revolution seems to more be about computer models being able to detect strong signals that people have been able to detect for a long time, like image recognition. This is great for automation. It's less clear to me how much it will improve detecting much noisier signals like propensity for loss.
I'm no luddite, but there is a danger that it will pick up signals from the training data which are artifacts of said data.

Having trained a neural net, I can show you that it has lift, but I can't explain how it is making its judgement. Maybe it just picked up for the training data that "we don't insure THOSE people." In my view, that should be a real issue for regulators.
"What do you mean I don't have the prerequisites for this class? I've failed it twice before!"

"I think that probably clarifies things pretty good by itself."

"I understand health care now especially very well."
Reply With Quote
Old 04-14-2019, 03:26 PM
therealsylvos's Avatar
therealsylvos therealsylvos is offline
Join Date: Mar 2014
Posts: 17,620

The article author gained a lot of credibility when they claimed sex was a prohibited rating variable.
Originally Posted by ao fan View Post
i like laying in the tube and pretending that i'm dead.
Reply With Quote
Old 04-14-2019, 03:27 PM
campbell's Avatar
campbell campbell is offline
Mary Pat Campbell
Join Date: Nov 2003
Location: NY
Studying for duolingo and coursera
Favorite beer: Murphy's Irish Stout
Posts: 89,021
Blog Entries: 6

It is in Europe.

LinkedIn Profile
Reply With Quote
Old 04-15-2019, 04:37 AM
GargoyleWaiting's Avatar
GargoyleWaiting GargoyleWaiting is offline
Join Date: Sep 2009
Favorite beer: the closest one
Posts: 7,210

A few thoughts from the UK life perspective...

The data bit is nothing new, it must be 10 or more years since I sat in my first conference session while someone told us the wonderful stuff they'll be able to do with big data. I'm not sure I understand where the AI bit comes in to it? Is it the fear that we'll give up oversight of pricing models and let computers decide?

That's starting to touch on what I think is the interesting bit. If we get more data we could (maybe, perhaps) refine our pricing down to a very precise level. And, assuming that were possible, where does that leave the industry? Do we just have a load of risks that are uninsurable, or at least can't be offered on affordable terms? Or do we have a duty to make sure people have access to the insurance they need - be that a moral issue, or a commercial one if we need the business volumes to sustain the industry?

We're starting to see this in the UK for car insurance - where young drivers are facing huge premiums and needing to get dashboard cameras and devices to measure their driving in order to get any cover. While that's manageable, and eventually they'll grow up and get appropriate premiums, it's a different issue for life insurance - if anything you become less insurable as you age. I can see using data insights to nudge policyholders towards better behaviour, but it does feel like there's a limit to how much we can subdivide risks and maintain an industry - either eveyone ends up with group cover, or the individual insurance ends up looking more like investment than protection cover.
Originally Posted by UFActuary View Post
But the mosquitoes in New Brunswick Bay of Fundy did mess with my understanding of some limited loss functions
Originally Posted by King of the North View Post
Excel gave me #VALUE.

Edit: Nevermind, I was linking a sumif and didn't open the linked spreadsheet. It is now giving me #N/A.
Reply With Quote
Old 04-15-2019, 08:31 AM
Vorian Atreides's Avatar
Vorian Atreides Vorian Atreides is offline
Wiki/Note Contributor
Join Date: Apr 2005
Location: As far as 3 cups of sugar will take you
Studying for ACAS
College: Hard Knocks
Favorite beer: Most German dark lagers
Posts: 65,252

Most mass media writers are going equate "machine learning" to A.I. since the latter literally incorporates the former.
I find your lack of faith disturbing

Why should I worry about dying? Itís not going to happen in my lifetime!

Freedom of speech is not a license to discourtesy

Reply With Quote
Old 04-15-2019, 11:25 AM
exponentc exponentc is offline
Join Date: Oct 2018
College: JHU
Posts: 52

Does this mean insurance companies are going to hire more data scientists/programmers to do pricing?
Reply With Quote
Old 04-15-2019, 11:11 PM
Arthur Itas's Avatar
Arthur Itas Arthur Itas is offline
Join Date: Feb 2004
Location: Atlanta
Studying for Prostate exam
College: Hard Knocks
Posts: 20,225

All your base are belong to us.
Reply With Quote

data science, predictive analytics

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

All times are GMT -4. The time now is 07:25 PM.

Powered by vBulletin®
Copyright ©2000 - 2019, Jelsoft Enterprises Ltd.
*PLEASE NOTE: Posts are not checked for accuracy, and do not
represent the views of the Actuarial Outpost or its sponsors.
Page generated in 0.16681 seconds with 11 queries