From theory to practice - page 342

[Violetta Novak]

Nikolay Demko:

By reading every 2nd, then 3rd, etc. every n-th tick you actually get a chart of the closing prices.

And the distributions from this chart I have already filled you in.

At the beginning you will get the central peak decreasing, it will start to blur closer to normal and then the distribution will become bimodal.

To understand the process you need to study it at the edges and the edge measures are that with n=1 you get near to a log-normal distribution and with increasing n closer to n=100 you have a bimodal distribution. This means that the distribution has always been bimodal, it's just that due to discreteness at small n it overlaps each other and the picture is not clear.

So your idea of a study is an invention of a bicycle.

Alexander_K2:

Yes, Maxim - the graphs only perform a transformation to the simplest flow.

Further transformation of input stream is needed only for neural networks, for forecasting. After all, that's the whole point - to work with known distributions. For the event stream - Erlang distribution, for quotes sitting in it - Gaussian distribution. Then all the mathematics from Kolmogorov's work will work. Without such a transformation, it won't.

I myself would switch to neural networks, but I don't have Vissim NeurlNet module and I'm too lazy to learn another system.

But I stand by my opinion - the problem of all participants of branch "Machine learning" is exactly in preparation of predictors and nothing else.

I am waiting for your breakthrough. Then I will connect to your trading signals.

---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------

For event stream - Erlang distribution.

https://habr.com/post/208684/ And this is a pointer.

[Алексей Тарабанов]

Renat Akhtyamov:
A_K2 wiped his posts and deleted his friends from his profile, and you guys are just pissed off...

Wiped out the posts. Killed my friends. I'm out of here.

[Violetta Novak]

Nikolay Demko:

By reading every 2nd, then 3rd, etc. every n-th tick you actually get a chart of the closing prices.

And the distributions from this chart I have already filled you in.

At the beginning you will get the central peak decreasing, it will start to blur closer to normal and then the distribution will become bimodal.

To understand the process you need to study it at the edges and the edge measures are that with n=1 you get near to a log-normal distribution and with increasing n closer to n=100 you have a bimodal distribution. This means that the distribution has always been bimodal, it's just that due to discreteness at small n it overlaps each other and the picture is not clear.

So your idea of a study is an invention of the bicycle.

I absolutely agree that the study has to start with the tails of the distribution. On higher TF a polymodal distribution is visible, as several peaks are formed.

The article here is a good one, and the system presented partially overlaps with Alexander's system.http://www.altertrader.com/publications01.html

[Renat Akhtyamov]

Wizard2018:

Well, to predict, not to predict, but on the Wiener process North Wind has"earned". Not by chance, that is guaranteed. / Algorithm is open, if you find a mistake, come to us.

https://forum.fxclub.org/threads/32942-prostye-nenuzhnye-veshhi

https://forum.fxclub.org/threads/32942-prostye-nenuzhnye-veshhi?p=600580&viewfull=1#post600580

https://forum.fxclub.org/threads/32942-prostye-nenuzhnye-veshhi?p=601363&viewfull=1#post601363

Here is a description of a sane person.

That is, a quasi-volumetric tick chart is sufficient, and time should be completely forgotten about

[secret]

Wizard2018:

Well, to predict or not to predict, but Northern Wind has managed to "make money" out of the Wiener process. Not by chance, i.e. guaranteed. The SSC and the SB are different. / Algorithm is open, if you find a mistake, come to us.

It is not different. MTB is a test model without regularities.

If someone has invented his/her own series, it is no longer CRS and it is not SB. I can artificially introduce regularities into SB, then it will be possible to earn on it, but it cannot be called SB.

SB does not contain patterns, that is its definition, it is constructed so as not to contain them. So "making money" on SB is like saying "2+2=5". That's not my opinion, it's a mathematical fact from textbooks. At least read the definitions on Wiki for an immersion into the subject.

And yes, the market is not a SB, even though it looks similar.

[Yuriy Asaulenko]

basilio:

It doesn't happen to be different. SB is a test model without a pattern.

If someone comes up with their own series, then it's no longer GSH and no longer SB. I can artificially introduce regularities into SB, then it will be possible to make money on it, but it cannot be called SB.

SB does not contain patterns, that is its definition, it is constructed so as not to contain them. So "making money" on SB is like saying "2+2=5". That's not my opinion, it's a mathematical fact from textbooks. At least read the definitions on Wiki for an immersion into the subject.

And yes, the market is not an SB, even though it looks like one.

Actually, you can make money on SB (not to be confused with a coin - although you can also make money on a coin, under certain conditions). It's quite obvious, but it's long, painful and lazy to explain it.

The market is of course not SB, but SB-like. In general, it is possible to make money in the market as in SB.

[secret]

SB is an integral of a coin) you can't make money under any circumstances, it's a mathematical fact from textbooks. It's like trying to disprove Pythagoras' theorem)

[Yuriy Asaulenko]

basilio:
SB is an integral of a coin) you can't make money under any conditions, it's a mathematical fact from textbooks. It's like trying to disprove Pythagoras' theorem)

)) Put SB in a confined space and you will have that opportunity.

[ILNUR777]

Yuriy Asaulenko:

)) Put SB in a closed space, and you will have such a possibility.

Here we go. I can't get past this repetitive idiocy. They start out like a tinhorn on a cooker. What if this, what if that. It's not wandering with demolition. What's that got to do with confined spaces.

Take an infinite number of such finite closed spaces and you end up with an equally infinite NOT closed space.

[Yuriy Asaulenko]

ILNUR777:
Here we go. Here I can't get past this repetitive idiocy. They start out like a tinhorn on a cooker. What if this, what if that. It's not rambling with demolition. What's that got to do with confined spaces.

Take an infinite number of such finite closed spaces and you will end up with the same infinite NOT closed space.

I see, you are all about the eternal and the infinite, but about the ideal).

Actually, in general case you won't even have a way to distinguish closed volume from infinite.