The Sultonov Regression Model (SRM) - claiming to be a mathematical model of the market. - page 45
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
If we solve the problem of an ellipse, which indicates a high correlation coefficient.
Let's calculate them for different periods (we took a period of 15), instrument types and historical data and find out their mutual correlation. In the meantime we can only state that they are not equally important in formation of the future candlestick.
Let's calculate them for different periods (now we took a period of 15), instrument types and historical data, and find out their mutual correlation. In the meantime, we can only state that they are not equally important in the formation of the future candlestick.
Sketched out a simple EA and with the available odds a constant lot "tested the formula" for the last year.
My terminal has not been saving stats for the last month, that's why pictures:
The second part of the test (along the time series) was more successful, and it is on this time interval that the sample of prices is located, on which the coefficients have been calculated.
You can say that "you can make a grandmother out of this grandfather" - if you try hard... :)
P.S. And if we delete orders with "hopeless prediction", it will be even prettier:
How complicated is the mathematical apparatus for calculating these coefficients?
I noted above that my proposed method does not present any difficulties - there are about 60 columns of simple calculations on exel, the program is ready to be coded on µl, I will pass it to my loyal programmers and soon we will make an indicator that can cover any number of bars. I am preparing an article with the basics of how to calculate the coefficients with all the formulas. Today I will send it to mcl4 or 5 for editing. It will depend on their willingness and efficiency when it comes out, but they need to like the topic first. Waiting for feedback from moderators and rosh.
I noted above that my proposed method does not present any difficulties - there are about 60 columns of simple calculations on exel, the program is ready to be coded on µl, I will pass it to my loyal programmers and soon we will make an indicator that can cover any number of bars. I'm preparing an article with the basics of how to calculate coefficients with all the formulas.
Sketched out a simple EA and with the available odds a constant lot "tested the formula" for the last year.
My terminal has not been saving stats for the last month, that's why pictures:
The second part of the test (along the time series) was more successful, and it is on this time interval that the sample of prices on which the coefficients were calculated is located.
You can say that "you can make a grandmother out of this grandfather" - if you try hard... :)
yosuf:
I am preparing an article with the basics of how to calculate the coefficients with all the formulas. I will send the article today to the editors at mcl4 or 5. It will depend on their willingness and promptness when it will come out, but first they need to like the topic. Waiting for feedback from the moderators and rosh.
Why tell the world about the grail!?
How complicated is the mathematical apparatus for calculating these coefficients?
It does not matter. The method of solving the equation does not affect the solution, if it exists and is singular, but if there are many acceptable solutions and Yusuf's is a local minimum, genetics or bees is better. For a head-on manual solution, just use a tester: the genetic algorithm will help you.
There is a danger of slipping into fitting - with so many varying parameters you need to take more examples to optimise.
One more remark - you don't know the sample period when the coefficients were searched. If this period coincided with the period of balance growth on the chart, then alas, everything is not so bright.
Thanks for the compliment, to be honest, I am a theorist who has solved a 200 -year-old problem since Kramer and Gauss, just finished, I think for Gauss, sooner or later must provide this method to the scientific community, so the private grail in Forex is not interesting for me, I do not need money in large quantities. But if traders will appreciate this indicator and will pay me a small, but stable percentage of their profits - I will not refuse. I am sorry for digressing from the subject of the discussion.
"I have erected a monument to myself..."