SERVING THE QUANTITATIVE FINANCE COMMUNITY

 
User avatar
Paul
Posts: 9597
Joined: July 20th, 2001, 3:28 pm

Re: DL and PDEs

September 3rd, 2019, 7:02 pm

This could be a situation where less is more. I.e. keep quants and their models out of it. So don’t have a model that makes silly assumptions about what is stable.
 
User avatar
Cuchulainn
Posts: 60256
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: DL and PDEs

September 3rd, 2019, 8:01 pm

What will we replace them by?
 
User avatar
ISayMoo
Posts: 2143
Joined: September 30th, 2015, 8:30 pm

Re: DL and PDEs

September 4th, 2019, 11:08 pm

There are no models without assumptions!
 
User avatar
Paul
Posts: 9597
Joined: July 20th, 2001, 3:28 pm

Re: DL and PDEs

September 5th, 2019, 1:08 am

I think you’ve just assumed adjectives don’t matter!
 
User avatar
ISayMoo
Posts: 2143
Joined: September 30th, 2015, 8:30 pm

Re: DL and PDEs

September 5th, 2019, 7:16 am

What are the not silly assumptions about what is stable?
 
User avatar
Cuchulainn
Posts: 60256
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: DL and PDEs

September 5th, 2019, 10:32 am

I think you’ve just assumed adjectives don’t matter!
Even worse, adjectives are dangerous (remember? Brutus is an honest man).

They've a temper, some of them--particularly verbs: they're the proudest--adjectives you can do anything with, but not verbs--however I can manage the whole lot of them!

Adjectives should be banned. For example you don't hear people saying Fantastic France or Brilliant Germany. Or analytical solution of American options, fast calibration etc. 
Last edited by Cuchulainn on September 5th, 2019, 11:02 am, edited 1 time in total.
 
User avatar
Cuchulainn
Posts: 60256
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: DL and PDEs

September 5th, 2019, 10:57 am

what is stable.
Many to choose from: [$]A[$], [$]A_0[$], [$] L[$], [$]A(\alpha)[$]
 
User avatar
Paul
Posts: 9597
Joined: July 20th, 2001, 3:28 pm

Re: DL and PDEs

September 5th, 2019, 2:14 pm

What are the not silly assumptions about what is stable?
Good question. Maybe there aren’t any.

The best finance assumption of the 20th century is that portfolios (stocks or option plus stock) have nicer properties than single instruments. Work with that but don’t go to the extremes of Markowitz and BSM.
 
User avatar
ISayMoo
Posts: 2143
Joined: September 30th, 2015, 8:30 pm

Re: DL and PDEs

September 5th, 2019, 7:39 pm

BSM < BDSM.
 
User avatar
Paul
Posts: 9597
Joined: July 20th, 2001, 3:28 pm

Re: DL and PDEs

September 5th, 2019, 8:27 pm

Derman? I am shocked!
 
User avatar
ISayMoo
Posts: 2143
Joined: September 30th, 2015, 8:30 pm

Re: DL and PDEs

September 5th, 2019, 8:51 pm

Close but no CIGAR!
 
User avatar
Cuchulainn
Posts: 60256
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: DL and PDEs

September 9th, 2019, 12:29 pm

Tensorflow quant library

https://github.com/google/tf-quant-finance

Any one used it?
 
User avatar
FaridMoussaoui
Posts: 446
Joined: June 20th, 2008, 10:05 am
Location: Genève, Genf, Ginevra, Geneva

Re: DL and PDEs

September 9th, 2019, 7:06 pm

It is in development (I would say in its enfancy) and expected to use Quantlib
 
User avatar
Cuchulainn
Posts: 60256
Joined: July 16th, 2004, 7:38 am
Location: Amsterdam
Contact:

Re: DL and PDEs

September 9th, 2019, 7:47 pm

enfant terrible?
Yet another ADI implementation. Google must have lots of resources.
Why not just stick to QL?
 
User avatar
ISayMoo
Posts: 2143
Joined: September 30th, 2015, 8:30 pm

Re: DL and PDEs

September 9th, 2019, 8:47 pm

Looks like someone's 20% project.

I prefer things like this: https://deepmind.com/blog/article/using-machine-learning-to-accelerate-ecological-research
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