Where is the line between fitting and actual patterns? - page 52

New comment
 
Roman.:


Guys, as for the classics of the genre, they say that the lower the TF, the higher the noise and I agree with that... They also think that movements on higher TFs are more 'predictable', whatever way you look at it... And in general, I think the answer to the starter's question has been found... Next steps...?

Unfortunately I do not know the truth, but it is possible that in the lower frames the movement obeys somewhat different models than in the higher, more complex - more difficult to predict at first glance (or simple methods).
 

granit77: Идет эксперимент. Лассо предложил банить на две недели всех, кто влезет с флудом в чужую тему (пресловутый 8-й пункт). Федор поймал его на флуде. Лассо с криком - "А я попадусь, стреляй меня, не жалей Чапаева!", согласился.

Now doing a fair bit of time until 11.02.

In fact, there was more - not just flubbing, but provocative and abusive behaviour. So you could have done without "point 8"...

 
-Aleksey-:
Unfortunately, the truth is unknown to me, but it is possible that on smaller frames the movement obeys somewhat different models than on older ones, more complex - harder to predict at first glance(or simple methods).


This is understandable. If you read the branch: 1st type, 2nd type, if you're into the subject, you seem to have made up your mind. The answer to the branch's question is here What next?

P.S. This question is not for you personally, but for all the members of the branch.

 
granit77:

There's an experiment going on. Lasso has offered to ban for a fortnight anyone who breaks into someone else's topic (the notorious 8th point). Fedor caught him flooding. Lasso, with a cry of "I'll get caught, shoot me, do not spare Chapaev!", agreed.

Now he is doing his time until 11.02.


That's how it is! So maybe the ban is ordered by the user's wish, a kind of stop-out level so that the Margin Call did not come? )))

SZS: Eh was not through my 54 posts and I'm screwed ;)

 
Mathemat:

In fact, there was more - not just flubbing, but provocative and abusive behaviour. So you could have done without "point 8"...


And that's understandable too. Just maybe, someone will have an impression of him as a kind of sufferer "for the truth in the womb" - just look at his snack for returning the nickname lasso, after a previous ban - when he spoke sp9, so it would be - And who is it ...? "Oh-oh-oh, it's the National Hero!!!" shh-shh-shh, just a whisper... :-)))
 
IgorM:

That's how it is! So maybe the ban should be ordered at the user's request, a kind of stop-out level so that the Margin Call does not come? )))
SZS: eh was not in 54 my posts and me to the top ;)

Do not laugh, there really were such requests. You should have read my personal list, you'd see a lot more. :))
 
Roman.:


It's understandable. If you read the thread: type 1, type 2, if you are on the subject, it seems to have been decided. The answer to the branch's question is here What next?

P.S. This question is not for you personally, but for all the members of the branch.

The link does not work. Further we can draw a logical conclusion that the fit is such a description of a pattern that does not (or cannot) describe the behavior of price on different frames by models of different complexity simultaneously. It turns out that if a model is not multiframe (it models only on one frame) - then it is a fit for this frame. So, the most non-fitted models are on minutes or multiframes. And ideally, multicurrency in crosses as well. Admittedly, it is probably already impossible to work with such volumes of data - one will need a super computer.
 
-Aleksey-:
Link does not work. Further, we can draw the logical conclusion that the fitting is such a description of a pattern that does not (or cannot) describe the behavior of price on different frames by models of different complexity at the same time. It turns out that if a model is not multiframe (it models only on one frame) - then it is a fit for this frame. So, the most non-fitted models are on minutes, or multiframes.

https://www.mql5.com/ru/forum/131308/page50 - next... what? (where to dig, according to the questions - answers of the thread)

P.S. Read it here.

PPS. I can't see or hear the starter itself...

PPS. With Lasso - got it.

 
Roman.:
https://www.mql5.com/ru/forum/131308/page50 - next... what?
Thanks, I read about the aftermath. Is this what you mean? What follows is one of the final conclusions concerning the topic of the branch: the fringe defines the ability or inability of a system (TS) to make consistent forecasts (or pattern descriptions) using information from different frames simultaneously, mainly from small ones, because they are more difficult to describe. But if the model on minutes is applied to a weekly trend - no resources are enough - we need some reasonable compromises.
 
-Aleksey-:
Thanks, I've read about the aftermath. Is that what you mean? Further, one of the final conclusions on the topic of the branch: the edge determines the ability or inability of the system (TS) to make consistent predictions (or pattern descriptions) based on information from different frames simultaneously, mostly from small ones, because they are more difficult to describe.

including, in general - completely about the whole page...
1...454647484950515253545556575859...64
New comment
Reason: