MACHINE period with minus value - page 28
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 Kalman thing is very clear here...
Thank you very much. This is, indeed, a whole NEXT...
I'll have a look...
Thank you very much. It's really a whole HUGE thing...
I'll have a look...
Privalov's thread on this is a little less than complete. IMHO, one of the most useful threads on the forum.
:)))) The only way to look into the future is to look into the past.
The problem is solvable.
Draw the MA with period x. We draw MA with period x and with a shift minus x. Then we need to complete the necessary tail in x bars of the shifted MA. The task is reduced to a pattern recognition problem: we look for on the whole (or a given) set of history data a maximally similar current segment of the last N bars (we only consider its slope (increment)) so that it lags behind the current one by more than x bars. Its continuation in x bars is the required tail. It will fit perfectly as the tangent line will maximally coincide. The only disadvantage is that this tail will "bounce" (change) all the time when the current situation changes (like ZigZag).
I will definitely implement it and post it in the indicators.
:)))) The only way to look into the future is to look into the past.
The problem is solvable.
Draw the MA with period x. Draw MA with period x and shift minus x. It only remains to complete drawing of a necessary tail in x bars of the shifted MA. The task comes down to a pattern recognition problem: we look for on the whole (or a given) set of history data a maximally similar interval of N bars (considering only the MA slope (increment)) so that it lags behind the current one by more than x bars. Its continuation in x bars is the required tail. It will fit perfectly as the tangent line will maximally coincide. The only disadvantage is that this tail will "bounce" (change) all the time when the current situation changes (like ZigZag).
I will definitely implement it and post it in the indicators.
:)))) The only way to look into the future is to look into the past.
The problem is solvable.
Draw the MA with period x. We draw MA with period x and with a shift minus x. Then we need to complete the necessary tail in x bars of the shifted MA. The task is reduced to a pattern recognition problem: we look for on the whole (or a given) set of history data a maximally similar current segment of the last N bars (we only consider its slope (increment)) so that it lags behind the current one by more than x bars. Its continuation in x bars is the required tail. It will fit perfectly as the tangent line will maximally coincide. The only disadvantage is that this tail will "bounce" (change) all the time when the current situation changes (like ZigZag).
I will definitely implement it and post it in the indicators.
If you are able to recognize images reliably enough, do you really need a mask?
And the recognition of images has to learn how to apply it for profitable trading.
And a positive answer to the question "will recognised patterns from the past work in the future" is not obvious.
If you can recognise images reliably enough, do you really need the waving machine itself?
And pattern recognition has to be learned how to apply it to profitable trading.
And a positive answer to the question "will recognised patterns from the past work in the future" is not obvious.
I agree.
Obviously there is no grail. But there is a mathematical statistics and probability theory :)
Grand merci to borilunad for the advice to check out wmlab's creation: https://www.mql5.com/ru/code/10755
I agree.
Clearly, there is no grail. There is, however, mathematical statistics and the theory of probability :)
Grand merci to borilunad for the tip to check out wmlab's creation: https://www.mql5.com/ru/code/10755
If you choose to thank me in French, then not grand merci, but merci beaucoup! If you decided to be saronic about the "creation",wmlab's attempt, then I only mentioned it because it's not a new solution, and I wasn't advising you anything at all. In general, I'm lukewarm about these prognosticating topics, another thing is the tendency, as they say, "don'tsit up against the wind". And if something has come out too late or too early or in the wrong direction, I blame myself above all and not the market. You have to be more cautious or courageous, the psychological factor is more important! However, an Expert Advisor, like a chess player's head or a chess computer, cannot absorb all the theory and practice, because they always face the opponent, who plays with known rules, which cannot be said about our "opponent. Therefore I agree with the important role of mathematical statistics and probability theory, but how to apply it, everyone comes to it independently, in his own way. All the templates, clues are excluded here. If my Expert Advisor even draws, I mean Breakeven, I feel satisfied.
I liked very much your design of channel indicator and it is interesting to learn about possible using channels in your own EA. Just for the experience and I can only use what I have built and tested.
Regardless, respectfully yours, Boris.
:)))) The only way to look into the future is to look into the past.
The problem is solvable.
Draw the MA with period x. We draw MA with period x and with a shift minus x. Then we need to complete the necessary tail in x bars of the shifted MA. The task is reduced to a pattern recognition problem: we look for a current segment of the last N bars on the whole (or a given) set of history data (we only consider its slope (increment)) so that it lags behind the current one by more than x bars. Its continuation in x bars is the required tail. It will fit perfectly as the tangent line will maximally coincide. The only disadvantage is that this tail will "bounce" (change) all the time when the current situation changes (like ZigZag).
I will definitely implement it and post it in the indicators.
You have a good spatial imagination. Try just imagining that you are chasing the price, chasing it, not knowing where it will go. You'll probably catch up :)
:-)
Beautiful! +
You could first try taking out a lightknuck at the beginning of a battle in Worldsoftanks with an artifact!
If it's ok, my nickname there is Fabio_1 - I'm downloading fireflies from France to WatChat... :-)