The Sultonov system indicator - page 88

 
Yousufkhodja Sultonov:
SLT_set_RF - test after optimization at RoboForex
SLT_set_from_RF - test at Alpari with settings from RoboForex
SLT_set_own - test after optimization at Alpari

Yusuf, apparently not only you, but many others do not know the significance of optimisation.

In fact, when you will get great results on some combination of input parameters, it means nothing.

Why?

Because if you have very many input parameters, you can make many million passes during optimization and choose the best combination of input parameters from them.

This result doesn't mean anything either, because probability is such a thing, that you can always find such combination of input parameters, at which the program bypasses all dangerous zones on historical data.

And it is enough to shift the values of all these "optimal" parameters a little to the left and to the right and you will see how stable it is on these changes.

If they will differ very much from each other, then such a strategy, bad.

I don't speak Russian, but I hope it's clear.

 
Renat Akhtyamov:
has the demo been loaded?

Let's launch after comprehensive testing and validation of the 'heavy artillery' option - the option with a period of 10....100......1000 bars of history.

 
Yousufkhodja Sultonov:

Let's launch after exhaustive vetting and testing of the "heavy artillery" option - the 10....100......1000 bars history perimeter option.

well, yeah, not bad.

It's good to have multiple options.

 
Yousufkhodja Sultonov:

Let's launch after exhaustive testing and verification of the "heavy artillery" option - the 10....100......1000 bars history perimeter option.

And this is where you can go into more detail. About bars 100, 1000

In your indicator

Virtual Level of Market Prices

which Trishkin wrote, you can change the number of bars in history.
?????
 
Renat Akhtyamov:

Yeah, not bad.

it is good to have a few options

On ticks - up to 1 million ticks, if there is such information in the depths of history, on the principle: 1 tick - 1 bar. I think the power and speed of the computer should be enough.

Let's try to cover all available history for all symbols, for all TFs from ticks, if available, to one-minute ones, up to one-month.

But I see that none of the programmers are in a hurry to help me in this ambitious task. Only two programmers from this list https://www.mql5.com/ru/forum/307935/page79#comment_11212381 are working hard.

 
Олег avtomat:

retard...

Gentlemen - dare I point out and express my opinion that there is no need to go off the deep end in terms of personal insults.... :-)

It's clear that the maths isn't good, but His Majesty - FUSSIS - rules!!!!

 
Yousufkhodja Sultonov:

On ticks - up to 1 million ticks, if there is such information in the depths of history, according to the principle: 1 tick - 1 bar. I think the power and speed of the computer should be enough.................

Geez, Yusuf - you seem to be going deep and far again, in terms of bending over and working off ALL and EVERYTHING!!!

Meanwhile, the power plant is not finished yet.... :-) It needs money...

I am not touching you in good faith.

...........

"But I see none of the programmers are in a hurry to help me with this daunting task. Only two programmers from this listhttps://www.mql5.com/ru/forum/307935/page79#comment_11212381are working hard."

Getting cheeky? No? JOB - don't rule?

Системный индикатор Султонова
Системный индикатор Султонова
  • 2019.04.03
  • www.mql5.com
Уважаемые форумчане, в качестве основы стратегии будущего индикатора рассмотрим и обсудим следующую гипотезу: Цена текущего бара зависит от 4-х зна...
 
MASTERXAYS:

And from here you can go into more detail. About bars 100, 1000

In your indicator

Virtual Level of Market Prices

which Trishkin wrote, you can change the number of bars per history.

There is no limitation on the number of historical data divisible by 9 in the quadrangle of the input data matrix. Now we have implemented the minimum possible square matrix M(5x5), where the number N(5x5) of input historical data involved is N(5x5)=5+5-1 =9 pieces in each matrix M(5x5) of each, unlimited, number of calculation cycle, as I showed earlierhttps://www.mql5.com/ru/forum/307935/page84#comment_11219540 . Now, we will implement a quadrilateral matrix M(5xm) consisting of m rows and 5 columns, in which the number of source data should be N(5xm)=5+m-1 pieces. This is how we will calculate the necessary quantities of historical source data. Therefore, I ask the programmers to make appropriate changes in the indicator settings with the possibility of changing the number of rows from m=5 to M>5. The maximum number of lines M is not limited and limited only by the total amount of available historical data for each, analyzed instrument.

As an example, I will give the calculation of the initial data with the matrix M(5x20) with the total number of the initial data N(5x20)=5+20-1=24, which is shown in the attachment.


Calculations were made using the following equations:



The verdict of indicator after processing of 20 bars: SELL, because a4 <1 and a0>0.

Системный индикатор Султонова
Системный индикатор Султонова
  • 2019.04.04
  • www.mql5.com
Уважаемые форумчане, в качестве основы стратегии будущего индикатора рассмотрим и обсудим следующую гипотезу: Цена текущего бара зависит от 4-х зна...
 
Roman Shiredchenko:

Gentlemen - I dare say and express my opinion that there is no need to go off on a spree in terms of personal insults.... :-)

It's clear that the maths is no good, but His Majesty - FUSSIS - rules!!!

it's not an insult.

it's a diagnosis.

 
Олег avtomat:

it's not an insult.

It's a diagnosis.

Forget it. This isn't the kind of thing where you get to compare notes.