Linear regression channel - page 8

 
Nikolai Semko:

Dimitri, one look at you and the first thought is how smart you are:



But why are you always talking such rubbish?

Is that it? Are you out of here? You don't have anything substantive to say?

 
Nikolai Semko:

I remembered the joke"Dad, who were you talking to just now".
But seriously, I am deeply convinced that parabolic regression rules. By virtue of its quadratic nature. Just as in the material world, the gravitational influence between two objects is inversely proportional to the square of the distance:


the same laws apply in the world of money.

:))) refers to autoregressive patterns that depend on previous prices, not time. I.e. input patterns are wavelets, output patterns are prices. Then the whole thing is generalized using Monte Carlo, for example. It will not work that fast :)

Simply, due to the non-stationarity of BP, the regression is quickly fouled for a while, usually
 
Dmitry Fedoseev:

And then, boom, marketing genius comes out of nowhere?

genius who? and why genius?

are your shoes tight again?
 
Maxim Dmitrievsky:

Genius of who? And why genius?

Are your shoes hurting again?

Just a stupid hit job?

 
Dmitry Fedoseev:

And you will not calculate RMS without a cycle. Without a cycle, you can calculate something similar to RMS, but not RMS. This has even been demonstrated here.

What do you want to bet on?
There is a BB.mql5 indicator in the MT5 standard examples.
If I fix it and the SCO is counted without a cycle for each successive bar and the channel is exactly the same as the original, what can I get from you?
I'll settle for anything.

 
Dmitry Fedoseev:

Just a dumb move?

I guess... what are we talking about? I'm talking about the writer and regression, and you're taking it personally again?

 
Nikolai Semko:

What do you want to bet?
There is a BB.mql5 indicator in MT5 standard examples.
If I fix it and the SKO is calculated without the cycle for each successive bar and it is exactly the same as the original one, what can I get from you?
I'll settle for anything.

But we won't see the code, will we? That's why I can say anything, too.

 
Maxim Dmitrievsky:

I guess... what are we talking about here? I'm talking about the writer and regression, and you're taking it out on me again?

So I'm talking about the other person too, and you're all about me. I fell for it once - I downloaded it and checked it out, but it turned out to be a lie.

 
Dmitry Fedoseev:

So I'm talking about the other person too, and you're all about me.

Ah... I get it... I haven't been following your Wall Street wars.

Well, it's pretty well done, what's not to mark. I'm in.
 
Nikolai Semko:

What do you want to bet?
There is a BB.mql5 indicator in the MT5 standard examples.
If I fix it and the SCO is counted without a cycle for each successive bar and the channel is exactly the same as the original, what can I get from you?
I'll settle for anything.

And were you willing to argue and prove obvious things. What for?