Using neural networks in trading - page 32

 
faa1947:

Errrrrr.... You're kind of baffling me - I thought that only stationary processes are dealt with in modern econometrics.

They say cognition can be historical or logical.

Historical. Until 1987 I completely agree with you. Stationarity. The dominance of Nobels with their efficient market and random wandering.

1987 - crisis in the markets. Thoughtful, but the Nobels continued to persevere. I think Black and Scholes set up a fund to demonstrate efficient markets. It burst by 1998.

After 1998, even more thought was given to the dubiousness of the idea of stationarity.

After the 2008 crisis, I don't come across any publications at all that refer to econometrics that consider stationary processes - only non-stationary ones. More theoretical publications are ready-made software code in R. A colleague called this code.

thanks for the historical reference, but we are talking about something else - non-stationary series lead to a stationary or pseudo-stationary form for analysis

P.S. How long have you been able to kamlava on R?

 
FAGOTT:

Thanks for the historical reference, but we are talking about something else - non-stationary series lead to a stationary or pseudo-stationary form for analysis

P.S. how much more can we camel on R?


I wish someone would show me why that damn R is so good. Somebody show me !!!! It's starting to feel like trolling.
 
FAGOTT:

You have been told - GARCH works with fixed rows.

FARIMA - don't know. Ready-made code - maybe. It works with stationary rows.

So?

I was taught otherwise. If there are references, please do. But your opinion contradicts everything I know. For example the wiki.
 
solar:

I wish someone would show me what makes this bloody R so good. Somebody at least show me !!!! It's starting to feel like trolling.
Nothing. You'll be less happy. But to each his own. Don't worry.
 
EconModel:
I was taught otherwise. If there are references, please do. But your opinion contradicts everything I know.

You have to start from the beginning.
 
EconModel:
Not at all. You will be less happy. And so to each his own. Do not worry.


I'm not worried.

Actually, I was wondering what you found so good there, but in view of your mysterious answers to sit for 5 years and so on, I got contradictory impressions.

 

What I don't understand is what successful, cool econometricians are doing on this godforsaken forum. There are no futures on MT4 and there cannot be.

 
TheXpert:

..... econometricians......fuchs no

where is the connection?

 
solar:


I'm not worried.

Actually, I was wondering what you found so good there, but in view of your mysterious answers to sit for 5 years and so on, I got contradictory impressions.


Handsome is beautiful.

My comrade-in-arms had this inscription on his shoulder... :-)

I think the subject is not covered, at least in terms of organization and preparation of input data to the network...

 
FAGOTT:
You have to start from the beginning

You have to start here. Start with a simple one.

Stationary series = Mo and variance is a constant. With ARCH the variance is not only not a constant, but also depends on previous values.

When building models, a check for ARCH residual from models is mandatory, because MOC cannot be applied in the presence of ARCH.