SERVING THE QUANTITATIVE FINANCE COMMUNITY

 
User avatar
Paul
Posts: 10775
Joined: July 20th, 2001, 3:28 pm

Re: How good or bad are the pandemic warning models?

April 5th, 2020, 9:15 pm

Boris hospital now 
Viral load? We might lose politicians, Royals, health workers, bus drivers, anyone who socialized a lot until recently, etc. Antisocials, like me, will be fine.
 
User avatar
katastrofa
Posts: 9213
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: How good or bad are the pandemic warning models?

April 5th, 2020, 9:23 pm

12. Easy...Because I’m very good at delegating!
It's called digital illiteracy.
Image
 
User avatar
Paul
Posts: 10775
Joined: July 20th, 2001, 3:28 pm

Re: How good or bad are the pandemic warning models?

April 5th, 2020, 9:30 pm

12. Why don't you ask uncle Google? ;-)
https://www.imperial.ac.uk/media/imperi ... 3-2020.pdf
That's dated 16 March, so they rely on China a lot (bad) but also South Korea (good). Have they updated it? I assume so.

As I understand it, South Korea has concentrated on testing and hunting down contacts, and has been extremely successful. 

So that's odd. The Imperial College model gets some insight from South Korea but not a strategy for coping? I'd say that the latter is more important!
 
User avatar
Cuchulainn
Posts: 62126
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: How good or bad are the pandemic warning models?

April 6th, 2020, 8:07 am

Neil Ferguson on TV last night said prediction very difficult.
 
User avatar
Collector
Topic Author
Posts: 4613
Joined: August 21st, 2001, 12:37 pm

Re: How good or bad are the pandemic warning models?

April 7th, 2020, 6:06 pm

Cured virus patients might have tested positive due to virus reactivation: KCDC

"More than 50 people who recovered after contracting COVID-19 have tested positive again, but the results might have been due to the reactivation of the novel coronavirus, health authorities here said Monday."
 
User avatar
Alan
Posts: 10167
Joined: December 19th, 2001, 4:01 am
Location: California
Contact:

Re: How good or bad are the pandemic warning models?

April 7th, 2020, 6:36 pm

Cured virus patients might have tested positive due to virus reactivation: KCDC

"More than 50 people who recovered after contracting COVID-19 have tested positive again, but the results might have been due to the reactivation of the novel coronavirus, health authorities here said Monday."
Interesting. But if the false negative rate is 30%, then there was about a 10% chance the "recovered" diagnosis (two tests) was also false. If so, then it didn't really reactivate. 
Last edited by Alan on April 7th, 2020, 6:45 pm, edited 1 time in total.
 
User avatar
Cuchulainn
Posts: 62126
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: How good or bad are the pandemic warning models?

April 7th, 2020, 6:41 pm

12. Why don't you ask uncle Google? ;-)
https://www.imperial.ac.uk/media/imperi ... 3-2020.pdf
That's dated 16 March, so they rely on China a lot (bad) but also South Korea (good). Have they updated it? I assume so.

As I understand it, South Korea has concentrated on testing and hunting down contacts, and has been extremely successful. 

So that's odd. The Imperial College model gets some insight from South Korea but not a strategy for coping? I'd say that the latter is more important!
The IC model doesn't model strategy methinks, Euler's method is great but it is a hammer.

Taiwan is not in WHO.
 
User avatar
Alan
Posts: 10167
Joined: December 19th, 2001, 4:01 am
Location: California
Contact:

Re: How good or bad are the pandemic warning models?

April 7th, 2020, 8:52 pm

The stock market is a pandemic warning model of sorts. But, remember the famous quip from Paul Samuelson (slightly adapted):
The stock market has predicted nine of the past five COVID-19 subsidences.
 
User avatar
trackstar
Posts: 27409
Joined: August 28th, 2008, 1:53 pm

Re: How good or bad are the pandemic warning models?

April 8th, 2020, 11:20 pm

Been studying some diverse perspectives on things and this is an interesting collection:

SFI (Santa Fe Institute) Insights on Covid-19

As some of you know, these tend to be complexity/complex adaptive systems guys and gals.

So, possibly useful for reading and discussion. Or discussion without reading. 

Or even loudly articulated and strongly held personal opinions without any thinking at all. 

Whatever you like - all good. 

Emitting both noise and signal mean you are still alive.

Enjoy! : )
 
User avatar
Cuchulainn
Posts: 62126
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: How good or bad are the pandemic warning models?

April 9th, 2020, 9:50 am

Been studying some diverse perspectives on things and this is an interesting collection:

SFI (Santa Fe Institute) Insights on Covid-19

As some of you know, these tend to be complexity/complex adaptive systems guys and gals.

So, possibly useful for reading and discussion. Or discussion without reading. 

Or even loudly articulated and strongly held personal opinions without any thinking at all. 

Whatever you like - all good. 

Emitting both noise and signal mean you are still alive.

Enjoy! : )
Some cliches..
It's probably not even wrong. All models are wrong, some are useful.

All I would like to see is the IC Ferguson model equations, data and parameters, use cases and conclusions. It is a modest proposal.

A leader of the calibre of Gordon Welcman (most people never heard of him) is needed to lead this project. It's not as if they are cash-strapped,
 
User avatar
katastrofa
Posts: 9213
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: How good or bad are the pandemic warning models?

April 9th, 2020, 12:08 pm

"All I would like to see is the IC Ferguson model equations, data and parameters, use cases and conclusions. It is a modest proposal."

The design and data description are actually available online (they re-calibrated the old model for influenza), as I wrote several times before. You need new spectacles ;-)
 
User avatar
Cuchulainn
Posts: 62126
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: How good or bad are the pandemic warning models?

April 9th, 2020, 2:05 pm

"All I would like to see is the IC Ferguson model equations, data and parameters, use cases and conclusions. It is a modest proposal."

The design and data description are actually available online (they re-calibrated the old model for influenza), as I wrote several times before. You need new spectacles ;-)
Nice try. You should never say that at a seminar/training.
Can you repost again, please? It got lost in the mêlée.
BTW you only mention design (whatever that's suppose to mean, it  is not 30 lines of Python) and "data", My question was more embracing.
 
User avatar
trackstar
Posts: 27409
Joined: August 28th, 2008, 1:53 pm

Re: How good or bad are the pandemic warning models?

April 9th, 2020, 3:21 pm

Looks like R0 might be more like 5.7 or so - not 2-3 as previously estimated.

Article in Forbes points to a study in the Emerging infectious Diseases journal.

The COVID-19 Coronavirus Disease May Be Twice As Contagious As We Thought - Forbes - April 7

Discusses herd immunity too, but posting a link here does not imply full endorsement of a line of thought.
 
User avatar
trackstar
Posts: 27409
Joined: August 28th, 2008, 1:53 pm

Re: How good or bad are the pandemic warning models?

April 9th, 2020, 3:26 pm

I'm not sure if you guys have already been through this assumption, but one of the fundamental assumptions of models created in general on this topic, and also including the Imperial College Model, is the length of time that social distancing needs to remain in place otherwise the virus will reemerge. To summarize, essentially it must remain in place until a vaccine or treatment is developed.

1. According to the Imperial College paper:

We show that intermittent social distancing – triggered by trends in disease surveillance – may allow interventions to be relaxed temporarily in relative short time windows, but measures will need to be reintroduced if or when case numbers rebound. Last, while experience in China and now South Korea show that suppression is possible in the short term, it remains to be seen whether it is possible long-term, and whether the social and economic costs of the interventions adopted thus far can be reduced.

Source: Impact of non-pharmaceutical interventions (NPIs) to reduce COVID- 19 mortality and healthcare demand 

2. MIT Technology Review - Estimate of 18 Months
3. UK Scientists - Estimate of 12 months

The other thing is how quickly will production recover after stay at home orders are lifted or only existing in part. In China, their recovery was rather quick. This leaves hope for what we will see in the United States when we begin walking a line between complete lock down and economic production.

1. Subway and traffic congestion have resumed in China
Image

2. Steel production has resumed in China
Image

3. Coal demand has begun to recover

Image
bumping this post up. from our new member "ikicker" last week.
 
User avatar
katastrofa
Posts: 9213
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: How good or bad are the pandemic warning models?

April 9th, 2020, 5:13 pm

"All I would like to see is the IC Ferguson model equations, data and parameters, use cases and conclusions. It is a modest proposal."

The design and data description are actually available online (they re-calibrated the old model for influenza), as I wrote several times before. You need new spectacles ;-)
Nice try. You should never say that at a seminar/training.
Can you repost again, please? It got lost in the mêlée.
BTW you only mention design (whatever that's suppose to mean, it  is not 30 lines of Python) and "data", My question was more embracing.
Model design (after John von Neumann) embraces such particulars. Just google Nature Fergusson influenza or his older paper on smallpox. For the record, I'm not defending those models. I think they are overly simplistic and without sufficient uncertainty assessment to be used to advise the decision makers. The whole country lockdown is not something I would call "strategy" anyway.
Who cares though? It's all politics and the same old story: the rich will gain, the poor will lose without even knowing - we will learn nothing.
https://www.youtube.com/watch?v=2xhmz7zBVTk
ABOUT WILMOTT

PW by JB

Wilmott.com has been "Serving the Quantitative Finance Community" since 2001. Continued...


Twitter LinkedIn Instagram

JOBS BOARD

JOBS BOARD

Looking for a quant job, risk, algo trading,...? Browse jobs here...


GZIP: On