Serving the Quantitative Finance Community

 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

November 17th, 2019, 11:54 am

This is a nice reply, JohnLeM. I have not seen any head-on comparison KS versus RKHS for 1-sample, 2-sample, independence tests etc. A cursory read of all this stuff is that the latter approach is more accurate (uniform convergence bounds, distribution-independent), robust, mathematically mature (Functional Analysis) and extendable. 
The DS statistic based on the empirical distribution function [$]F_n[$] comes across as rough-and-ready from a mathematical and numerical point of view (I am not a specialist on KS).The kernel method on the other hand (e.g. Gretton et al, equation (4)) allows us to define the MMD statistic ("MMD >> DS"). Indeed, it seems to work with low sample sizes. MMD(p,q) == 0 iff p == q. And it works on structured data.

A tongue-in-cheek remarks: "statistics don't do metrics, norms, Hilbert space, Cauchy-like sequences". I might be very wrong.

JohnLeM has done kernel methods for PDE.
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

December 17th, 2019, 12:15 pm

“People worry that computers will get too smart and take over the world, but the real problem is that they’re too stupid and they’ve already taken over the world.”

Pedro Domingos, cited in Artificial Intelligence: A Guide for Thinking Humans by Melanie Mitchell
 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

December 17th, 2019, 1:22 pm

I'm always in favour of having more thinking humans. Should we not train humans instead?

The above bland statement is soo banal. It's almost as bad as the waffle on negative probability. 

// I have a book by Mitchell on GAs.
 
User avatar
katastrofa
Posts: 7440
Joined: August 16th, 2007, 5:36 am
Location: Alpha Centauri

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

December 17th, 2019, 1:39 pm

Machines are better and cheaper.
 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

December 17th, 2019, 7:53 pm

I must say I have a bit of nostalgia for 60 bit supercomputers. I used it to simulate city heating Waterhammer (1st order pde (6), (7)) for The Hague a while back. Punch cards, real line printers, Fortran 66(IV).

It might come back again seeing that gas, wood and oil heating will be banned soon. Now what did I do with those punch cards.


https://www.win.tue.nl/analysis/reports/rana04-34.pdf

Image
 
User avatar
JohnLeM
Topic Author
Posts: 379
Joined: September 16th, 2008, 7:15 pm

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

December 18th, 2019, 5:49 pm

I must say I have a bit of nostalgia for 60 bit supercomputers. I used it to simulate city heating Waterhammer (1st order pde (6), (7)) for The Hague a while back. Punch cards, real line printers, Fortran 66(IV).

It might come back again seeing that gas, wood and oil heating will be banned soon. Now what did I do with those punch cards.


https://www.win.tue.nl/analysis/reports/rana04-34.pdf

Image
I want that laptop for christmas !
 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

December 18th, 2019, 8:56 pm

And the crash barriers? They're extra.
 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

April 19th, 2023, 1:01 pm

Time to move on from ANN drama?

This paper is their 10th attempt..


https://www.researchgate.net/publicatio ... _Equations

Meshuggah

Insanity Is Doing the Same Thing Over and Over Again and Expecting Different Results
 
User avatar
JohnLeM
Topic Author
Posts: 379
Joined: September 16th, 2008, 7:15 pm

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

April 19th, 2023, 1:10 pm

Time to move on from ANN drama?

This paper is their 10th attempt..


https://www.researchgate.net/publicatio ... _Equations

Meshuggah

Insanity Is Doing the Same Thing Over and Over Again and Expecting Different Results
Yep, it is really painful to read that again and again :/
 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

April 19th, 2023, 6:49 pm

What were the top FIVE incorrect implicit assumptions?

#1 UAT == Holy Grenade of Antioch.

Image
 
User avatar
Cuchulainn
Posts: 20253
Joined: July 16th, 2004, 7:38 am
Location: 20, 000

Re: Are Artificial Intelligence methods (AKA Neural Networks) for PDEs about to rediscover the wheel ?

April 19th, 2023, 6:57 pm

Basing approximation on Measure Theory and Legesgue Integral is wakky.

viewtopic.php?f=34&t=100821