Machine learning in trading: theory, models, practice and algo-trading - page 847
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
Whatever way you look at it, analyzing the behavior of the price as a stationary process or picking stationary noise-like components out of it is fundamentally illiterate and even wild from a mathematical point of view.
There has long been a mathematical basis for the analysis of nonstationary processes - it's another matter how to adapt it in practice.
Here is an example of analysis of unsteady BP using the well-known Viterbi algorithm.
Is there a full article or similar?
Is there a full article or similar?
well there's a link to the thesis...
Well, there's a link to the thesis...
10 pages only
10 pages only
there they ask to buy the full text of the dissertation... But I think you can order it from the library and there is a lot of stuff on this topic...
I'll add it to the TS, what :) no way, to buy all kinds of junk
it most likely describes signal filtering
I'll add it to the TS :) no way, to buy all sorts of junk
It's more likely to describe signal filtering
Rather the analysis of internal latent states of unsteady VR and their qualitative recognition from the mathematical point of view.
Signal filtering is a particular primitive practical case.
In any case the filtering of trading signals looks more competent and reasonable than filtering noise and catching fleas in it. :)
In any case the filtering of trade signals looks more intelligent and reasonable than filtering noise and catching fleas in it. :)
Finally - because I was called an earthworm for this approach! True, I use it not in NS, but in other ATSs.
Whatever way you look at it, analyzing the behavior of the price as a stationary process or picking stationary noise-like components out of it is fundamentally illiterate and even wild from a mathematical point of view.
There has long been a mathematical basis for the analysis of nonstationary processes - it's another matter how to adapt it in practice.
Here is an example of the analysis of nonstationary VR using the well-known Viterbi algorithm.
In general, it's extremely illiterate to look at the market as a non-stationary time series. It's a MARKET, not the parameters of some unit, where it works the same way from year to year. Take a broader look at the market. Take a class. Try to pass the FFMS certificate. At least read questions from there to understand what the market is.
As any researcher. The better a specialist in the IO should study the subject area and the better he does it, without rose-colored glasses and own silly assumptions, but to trust the facts describing and affecting the market. The sooner he will make a really good TS...
we just need to get the desired row directly in MT5 through custom symbols - this will eliminate the problems with packages and third-party software, it will be easy to get rows for any symbol
At the same time this would be a great example of how to create a symbol with an arbitrary distribution based on another
Should help.
That should help.
Good example, thanks ) I'll post the results when I do