Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Posts: 65000
Joined: July 16th, 2004, 7:38 am
Location: Ganymede

Re: Models for Covid-19 - analytics

July 26th, 2020, 12:53 pm

Another perspective on the current COVID situation in the US
(source: cdc)

WeeklyExcessDeaths.png
Does it matter? The US and UK approaches have been dismal failures.
65000
 
User avatar
bearish
Posts: 6704
Joined: February 3rd, 2011, 2:19 pm

Re: Models for Covid-19 - analytics

July 26th, 2020, 1:23 pm

Another perspective on the current COVID situation in the US
(source: cdc)

WeeklyExcessDeaths.png
Does it matter? The US and UK approaches have been dismal failures.

Ummm - I think you are making the assumption that keeping the number of deaths low was a major objective.
 
User avatar
Cuchulainn
Posts: 65000
Joined: July 16th, 2004, 7:38 am
Location: Ganymede

Re: Models for Covid-19 - analytics

July 26th, 2020, 5:23 pm

Another perspective on the current COVID situation in the US
(source: cdc)

WeeklyExcessDeaths.png
Does it matter? The US and UK approaches have been dismal failures.

Ummm - I think you are making the assumption that keeping the number of deaths low was a major objective.
Are you trying to talk me into a corner?
Check out my recent comments on COVID-19 thread.
65000
 
User avatar
bearish
Posts: 6704
Joined: February 3rd, 2011, 2:19 pm

Re: Models for Covid-19 - analytics

July 26th, 2020, 7:38 pm

For the UK case I’m willing to apply the old adage that one shouldn’t attribute to malice that which is adequately explained by incompetence. In the US, there is plenty of malice at work at the highest level of government. As well as oodles of incompetence. They are not mutually exclusive, after all.
 
User avatar
Cuchulainn
Posts: 65000
Joined: July 16th, 2004, 7:38 am
Location: Ganymede

Re: Models for Covid-19 - analytics

July 27th, 2020, 10:05 am

Another perspective on the current COVID situation in the US
(source: cdc)

WeeklyExcessDeaths.png
Does it matter? The US and UK approaches have been dismal failures.

Ummm - I think you are making the assumption that keeping the number of deaths low was a major objective.
Looking back, why almost empty hospitals in Manhattan and hospitals in Bronx bursting at the seams?
65000
 
User avatar
katastrofa
Posts: 10275
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Models for Covid-19 - analytics

July 27th, 2020, 10:59 am

Image
 
User avatar
Cuchulainn
Posts: 65000
Joined: July 16th, 2004, 7:38 am
Location: Ganymede

Re: Models for Covid-19 - analytics

July 27th, 2020, 11:08 am

I'm in top-left hand corner (0,0) and (1,0). I am however, considering (2,0).

I have an appointment with Junior-the-barber today. Have to look spik-n-span for my satellite speech for Thalesians 5 August.

Image
65000
 
User avatar
Cuchulainn
Posts: 65000
Joined: July 16th, 2004, 7:38 am
Location: Ganymede

Re: Models for Covid-19 - analytics

February 14th, 2021, 11:56 am

6 out of 10 people who have died from COVID-19 are disabled

https://www.health.org.uk/news-and-comm ... e-disabled
65000
 
User avatar
bearish
Posts: 6704
Joined: February 3rd, 2011, 2:19 pm

Re: Models for Covid-19 - analytics

February 14th, 2021, 3:16 pm

6 out of 10 people who have died from COVID-19 are disabled

https://www.health.org.uk/news-and-comm ... e-disabled

Something is deeply wrong with the grammar here. Surely, 10 out 10 people who have died from Covid-19 (or, for that matter, anything else) are disabled.
 
User avatar
Cuchulainn
Posts: 65000
Joined: July 16th, 2004, 7:38 am
Location: Ganymede

Re: Models for Covid-19 - analytics

February 14th, 2021, 5:43 pm

let me get back to you on that one tomorrow. Our proof readers have a day off.
65000
 
User avatar
Paul
Topic Author
Posts: 11442
Joined: July 20th, 2001, 3:28 pm

Re: Models for Covid-19 - analytics

August 15th, 2021, 1:22 am

Can someone (and by someone I obviously mean kat!) do the maths of probability of having covid given one positive test and one negative, taken simultaneously. Assume different parameters for the two tests (i.e. false pos, neg, etc.) to model different types of tests. And if you have these parameters for the different types of tests that would be great too!

Thanks in advance!
 
User avatar
ikicker
Posts: 29
Joined: June 8th, 2011, 12:19 am

Re: Models for Covid-19 - analytics

September 16th, 2021, 6:49 pm

In step with Paul’s test regarding PCR tests, the question has come up as to whether not vaccination increases or reduces the number of variants as well on LinkedIn. Hear me out…

1. One guy cites polio with a sample of 1 - a virus that was eliminated. However, corona viruses are resilient and have been around a long time. Corona virus and polio are very different.

2. If you look at a simple tree graph of the number of variants before and after the intervention (event) of the vaccine, the number of variants goes way up afterwards. However, I have a problem with this approach, despite my initiation that it is accurate (there is also at least one study that shows this in another virus). The issue is that in theory the number of variants is multiplicative or exponential in some manner. There is still the doubt that what I could be seeing is non-linearity.

I have a simple idea of how to test this using a Markov-model and a monte-Carlo simulation. However, what are the states and probabilities? When you move through different states, there is always a given probability that the virus will mutate. I could simply emulate it to come up with a similar tree once I have those probabilities, then I could compare reality to the baseline, and compute the variance by treating it as a mean.

Thoughts?
----
Undergraduate: accounting, finance, information systems; Graduate: MBA/finance; Graduate certificates: data science, applied statistics, advanced valuation; PhD candidate - data science

Blog: Www.ThinkerTinkerSolutions.com
 
User avatar
ikicker
Posts: 29
Joined: June 8th, 2011, 12:19 am

Re: Models for Covid-19 - analytics

September 16th, 2021, 6:54 pm

Can someone (and by someone I obviously mean kat!) do the maths of probability of having covid given one positive test and one negative, taken simultaneously. Assume different parameters for the two tests (i.e. false pos, neg, etc.) to model different types of tests. And if you have these parameters for the different types of tests that would be great too!

Thanks in advance!


You need a confusion matrix. This study has one, but of course the probabilities of being infected and in the hospital are different from in the general population.

Image
Study: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8080131/
----
Undergraduate: accounting, finance, information systems; Graduate: MBA/finance; Graduate certificates: data science, applied statistics, advanced valuation; PhD candidate - data science

Blog: Www.ThinkerTinkerSolutions.com
 
User avatar
ikicker
Posts: 29
Joined: June 8th, 2011, 12:19 am

Re: Models for Covid-19 - analytics

September 19th, 2021, 2:14 am

In step with Paul’s test regarding PCR tests, the question has come up as to whether not vaccination increases or reduces the number of variants as well on LinkedIn. Hear me out…

1. One guy cites polio with a sample of 1 - a virus that was eliminated. However, corona viruses are resilient and have been around a long time. Corona virus and polio are very different.

2. If you look at a simple tree graph of the number of variants before and after the intervention (event) of the vaccine, the number of variants goes way up afterwards. However, I have a problem with this approach, despite my initiation that it is accurate (there is also at least one study that shows this in another virus). The issue is that in theory the number of variants is multiplicative or exponential in some manner. There is still the doubt that what I could be seeing is non-linearity.

I have a simple idea of how to test this using a Markov-model and a monte-Carlo simulation. However, what are the states and probabilities? When you move through different states, there is always a given probability that the virus will mutate. I could simply emulate it to come up with a similar tree once I have those probabilities, then I could compare reality to the baseline, and compute the variance by treating it as a mean.

Thoughts?
The chart from the second half of the presentation would be what I am describing at 2:35.
----
Undergraduate: accounting, finance, information systems; Graduate: MBA/finance; Graduate certificates: data science, applied statistics, advanced valuation; PhD candidate - data science

Blog: Www.ThinkerTinkerSolutions.com