Regression equation - page 6
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
Are there no software implementations in the maths packages?
The next question will be "where are the screenshots" :-).
Look on Wikipedia, Quantile regression, there are links to statpacks.
or the link above the text. But there's practically nothing to be found in russian.
I sort of decided to slowly take up the program, I repeat the question -
Who knows where to find an implementation of linear programming, simplex at the very least, but better this or that??? Maybe someone has friends/acquaintances in the universities dabble :) Myself just terrible how lazy to dig:)
More specifically, I'm particularly interested in multivariate regression. Looking at options for solving nonlinear regression is also interesting. I have not found any algorithms to solve the multivariate regression in MQL. If you provide me with links and indicators (if you're not too lazy of course), it will be great!
Due to my lack of education three hours ago I still didn't know what regression, MNA and normal distribution are...
Multivariate linear regression in MQL can be seen here. However, it seems to be more advanced than multivariate linear regression (which, like non-linear regression, only requires solving a system of diff. equations (partial derivatives of the target function are zero)).
If I understand MNC correctly, it is simply minimizing the target function, which is the variance. The target function, of course, can also be defined differently. For example, not the sum of squares of variance, but the sum of absolute values. I am not yet familiar with the efficiency analysis of the different target functions.
Who knows where to find an implementation of linear programming, simplex at the very least, but better this or that? Maybe some friends/acquaintances in universities dabble:) Myself just terrible how lazy to dig:).
....
Who knows where to find an implementation of linear programming, simplex at the very least, but better this or that? Maybe some friends/acquaintances in universities dabble:) Myself just terrible how lazy to dig:)
Below are references to examples of implementation of numerical methods of unconditional minimization, which are simple, clear and clear enough to be immediately implemented in MQL:
Matcad is OK. But, since it is limited to a linear problem, it is apparently a simplex. I foresee problems with the complexity of the enumeration.
Regarding descent - will it work for non-smooth functions?
Matcad is OK. But since it is limited to a linear problem, it is apparently a simplex. I foresee problems with the complexity of the search.
Regarding descent - will it work for non-smooth functions?
I am a practitioner, not a theorist. I don't know how to solve problems in general terms. The conditions of the problem?
Come to a clear formalization of the target function, then it will be easier to look for a suitable working method of finding a solution.
I am a practitioner, not a theorist. I don't know how to solve problems in general terms. The conditions of the problem?
If you get to a clear formalization of the target function, then it will be easier to find a suitable working method for finding a solution.
Everything is already formalized, read the link, the one in Russian (the first one on page 3). Quantile regression problem is reduced to linear programming problem: find minimum of linear function under linear constraints.
I was thinking here, gradient descent will work worse than simplex-method, since grad-t is more general. All other things being equal, there are knowingly no fewer iterations.
In principle the article gives a hint as to how to reduce the number of iterations. So I'll probably write an "optimized" simplex for now. If I reach a computational limit, I'll think further: )))))