From theory to practice - page 1480

New comment
 
khorosh:

He has clearly told you that so far it is not working out, what is there to watch. It is not an easy task to catch zigzag extremes and not everyone can do it).

It's just not clear to me what he told you.
 
Martin_Apis_Bot Cheguevara:

it would be interesting to look at your work.

I wonder how it deals with time-suppressing flats and time-suppressing distance and minimum realisation time trends.

The easiest thing to do is to Skype and see on screen what's already functioning and what the little problem is. There is nothing wrong with the Zigzag
 
aleger:
The easiest thing to do is to Skype and see on the screen what is already functioning and what the little problem is. There is nothing wrong with Zigzag
When it comes to forex, problems are never small...
 
Martin_Apis_Bot Cheguevara:
When it comes to forex, problems are never small...
It depends...
 
aleger:
It depends...
Well then I guess what you have:
1. The problem of insufficient profits.
2. Works from time to time

In 1 problem there is no problem, increase the lots and the profit will grow.
In 2, the system itself is the problem.
 
Alexander_K:

So that sufferers don't think they have been forgotten.

Here is the book (see attached file). The Grail is there.

Thank you, Alexander, for a useful source. Though it is not about forex, (MIPT Methodological Materials. Popov P.V. Diffusion). It reveals patterns of diffusion processes in a concise manner. I like it because even at the content level


already, as a basis, draws attention to the square root law, which is akin to diffusion and forex, about which I have already written here several times:

https://www.mql5.com/ru/forum/193378/page16#comment_5116118
https://www.mql5.com/ru/forum/221552/page73#comment_6203173
https://www.mql5.com/ru/forum/221552/page77#comment_6208896
https://www.mql5.com/ru/forum/221552/page123#comment_6306015
https://www.mql5.com/ru/forum/221552/page430#comment_7945889
https://www.mql5.com/ru/forum/220237/page9#comment_6129706
And you, as far as I understand, use it when calculating scatter characteristics. I only want to support your movement to take into account, besides properties of distributions, also properties of nested periods (I guess that they are analogs of momentum relaxation time from the book). This is the right road to the analysis of sequences.

 
Vladimir:


Not at all, Vladimir. Glad you're not leaving the forum, as the presence and support of veterans is very important in any endeavour.

 
Martin_Apis_Bot Cheguevara:
Well then I guess that's your problem:
1. The problem of insufficient profits.
2. Works from time to time

In 1 problem there is no problem increase lots and profit will grow.
In 2 the system itself is the problem.
There is no "insufficient profit" problem at all - the work can be performed trend-wise and virtually continuously on the basis of all of the current and partially previous volatility actually occurring during the day, because it takes into account changes in quotations not only in the visualised but also in the hidden part of the market.
The work can be done trend-wise and actually continuously based on all the real current and part of the previous day volatility, because it not only takes into account the visualized but also the hidden part of the
of the current as well as previous and pre-existing trends.
 
Vladimir:

Thank you, Alexander, for the useful source. Though it is not about forex, (MIPT Methodological Materials. Popov P.V. Diffusion). It reveals patterns of diffusion processes in a concise way. I like it because even at the content level


already, as a basis, draws attention to the square root law, which relates diffusion to forex, as I have already written here several times:

https://www.mql5.com/ru/forum/193378/page16#comment_5116118
https://www.mql5.com/ru/forum/221552/page73#comment_6203173
https://www.mql5.com/ru/forum/221552/page77#comment_6208896
https://www.mql5.com/ru/forum/221552/page123#comment_6306015
https://www.mql5.com/ru/forum/221552/page430#comment_7945889
https://www.mql5.com/ru/forum/220237/page9#comment_6129706
And you, as I understand it, use it in calculating the characteristics of diffusion. I just want to support your movement to take into account, besides properties of distributions, also properties of nested periods (I guess that they are analogs of momentum relaxation time from the book). This is a sure road to sequence analysis.

1) Before drawing histograms on a sample, make sure that this sample is taken for a set of random variables satisfying the conditions of the Glivenko-Kantelli theorem. For forex, they are usually not fulfilled due to marked non-stationarity and dependence of the increments.

2) The analysis should be carried out at times of the order of the average transaction time. Analysis at much longer times makes no practical sense.

3) An arbitrarily taken sequence with a unit probability is not algorithmic and, therefore, not predictable accurately. The question is how its prediction errors are interspersed in time. If they are mixed evenly, that is relatively good. But in forex, they obviously have a tendency to cluster (a series of errors in the bad direction after a series of errors in the good direction), and this can easily destroy even a potentially profitable system. This phenomenon is easily masked by analysis on huge time intervals.

 
Renat Akhtyamov:
Yes, it's not the first time I've said the same thing since I've been on the subject for 4 years.
4 years down the road? :)
1...147314741475147614771478147914801481148214831484148514861487...1981
New comment
Reason: