Machine learning in trading: theory, models, practice and algo-trading - page 3365
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
The idea is to find more stable for profit parameters that regulate the optimised parameters. In addition, there is a possible feedback of influence on the parameters. But this is not an end in itself, but as a branch of research.
Do you propose to put f( X ) instead of X everywhere in TC? And instead of a linear f-function substitute something else, investigating how it affects the final result?
Of course, you can do any manipulations, but we were talking about something different from the beginning.
Are you suggesting to put f( X ) instead of X everywhere in TC? And instead of a linear f-function substitute something else, investigating how it affects the final result?
Of course, you can do any manipulations, but we were talking about something different from the beginning.
Yes, but given the power today, if only consciously, in some areas intuitively).
Today's paradigm does not allow it.)
Are you suggesting to put f( X ) instead of X everywhere in TC? And instead of a linear f-function substitute something else, investigating how it affects the final result?
Probably realised what concept is meant. By way of example.
Let's say there is a channel TS, where the trade goes inwards from the borders. And here I set the parameter to be optimised, which is a coefficient of the channel size (width).
The classical way is fine. You optimise it and see how it affects you.
According to the proposed concept, you cannot do it this way, because this parameter does not depend on the initial data (quotes history). In the proposed terminology, it is a "constant" optimised parameter.
Even if this parameter affects some polynomial, it is also a "constant" parameter, because there is no dependence on the CVR.
That's an interesting thought. Thanks.
Perhaps realised what concept is meant. By way of example.
Let's assume that there is a channel TS, where trading goes from the borders inwards. And here I set the parameter to be optimised, which is the coefficient of the channel size (width).
The classical way is fine. You optimise it and see how it affects you.
According to the proposed concept, you cannot do it this way, because this parameter does not depend on the initial data (quote history). In the proposed terminology, it is a "constant" optimised parameter.
Even if this parameter affects some polynomial, it is also a "constant" parameter, because there is no dependence on the VDC.
That's an interesting thought. Thank you.
I don't understand anything.
I don't get it.
Valery explains it better)
good book
Machine learning for factor investing
What does the MO do? Essentially picks a formula and variables to it from 100500 variables; and we can see from the OOS that we need to retrain sometimes. The chosen formula and the 3 parameters to it will most likely need to be changed too.
If it were some formula describing some law of physics/mathematics, it would be correct to use it. But I am afraid that in markets such invariable laws do not apply. You can pick up something experimentally, but you will pick it up on some separate part of history, and outside of it the picked up formula can work badly. And every week you will have to select not one n, but 3 parameters: x,y,z.