neural network and inputs - page 20
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
Same as for "non-fractals" ".... that is composed of several parts, each of which is similar to the whole figure." ©
This is what it literally means. A pattern consists of "small" patterns, and those in turn consist of even smaller ones, and so on down to the ticks, and up to infinity. But the structure of both "big" and "small" patterns never changes. What changes is the volatility (size) of the pattern.
I gave a link to Spider, there are some pictures with such patterns and how they are arranged in relation to the market.
Another link: It looks similar to the first picture in the article.
https://ru.wikipedia.org/wiki/%CA%F0%E8%E2%E0%FF_%CA%EE%F5%E0
It should be taken into account that the volatility of the pattern changes and also when searching for timeframes, there will be strong distortions due to the fact that the internal time and the astronomical time do not coincide. The process can run at different speeds and does not have to match the timeframes. And it doesn't, very rarely 'hitting' the timeframes.
Well, 28 again...
Fractals are geometric shapes. They can only have two dimensions, they have no third - time. If there are distortions, they are no longer fractals.
If a figure appears on 1H and when decreasing the dimensionality ALL 'that figure does not consist of similar, but smaller ones, it is not fractal
Stells:
about the similarity ....
I sometimes try to divide the movement into waves, and let's say on m5 I get a movement with an extended fifth wave,
and I kind of expect an elongated fifth wave on higher TFs.
Can it be called a fractal ? (movement with lengthened fifth wave)
You do break into waves by extrema (how else would you do it), and you filter the appropriate extrema somehow and look for them on a particular M5 timeframe or any other, it doesn't matter.
But the market does not live by your watch, it has its own, internal time, because of this there are a lot of problems. "There will be a lot of distortion because the internal time does not coincide with the astronomical time. The process can run at different speeds and does not have to match the timeframes. " (с)
When the process will "hit" the timeframe by accident, you select the right extremums to build a pattern, then sooner or later the process is bound to speed up or slow down and that's it, it's over. It works, it doesn't work, etc.
On the X-axis it looks like this:
I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I - astronomical time
I I I I I I IIIIIII I I I I I I I I I I I I I I I I - internal time
Well, 28 again...
Fractals are geometric shapes. They can only have two dimensions, they have no third - time. If there are distortions, they are no longer fractals.
If a figure appears on 1H and when decreasing the dimensionality ALL 'that figure is not composed of similar, but smaller ones, it is not a fractal
A pattern (like a fractal) is a geometric figure, but we construct it in price-time coordinates.
Yes? And who wrote that the patterns/rules work, then they don't and suddenly they disappear? It is no longer interesting how to avoid this and avoid guessing: will it work or won't it work?
To make NS work, you always have to define the concepts. A bad term, but it is a fact.
The machine has to understand what you want it to do. If the grid works and it doesn't, then you haven't defined the concepts.
And then dance around as best you can. That's all I'll say.
Any fractal is a pattern. You wrote, a few pages ago, that it is not possible to have patterns on history that are guaranteed (or highly likely) to work in the future. You define what you have there that disappears and what you don't.
All trolling, no specifics.
)))) If you mean that a fractal is a pattern, everything is clear )))) How could you understand what I wrote in such a perverted way?
In general, what I mean is that it is possible to earn on patterns (patterns are a special case). But over time, these patterns may not disappear, and the opportunity to earn on them with the help of TS, which uses these patterns, disappears.
Why, you ask? Because financial markets, in particular Forex, are not stationary.
What is non-stationary? Non-stationary is that the rules in the financial markets change, ie if before a certain movement, which catches your TS, was a pattern or series of patterns (regularity), then over time, this pattern may be in front of the opposite movement and your TS accordingly ceases to profit and starts to lose. And before the movement, which your TS catches, another pattern appears (regularity), which your TS does not know.
Once again, to be specific - it is not patterns (a special case - a pattern or a series of patterns) that disappear, but the opportunity to earn money with a particular TS, using these patterns, disappears.
But there are millions of patterns (a special case of a fractal) which exist in financial markets (Forex) and which are eternal, and never disappear )))). Three white candles, three black candles, black-white-black candles, etc. But there is no way to make money from them.
So that's what you're being asked - what are the patterns on which you can make money forever. Fractals? Well then show me how..... )))
Just two requirements:
1.Repeatability
2.self-similarity.
You are diving into the maze of mathematics, you forget about the third requirement - the possibility of earning, ie earning a profit, but in the best case, getting equal equity, which goes smoothly into the sky)))
In fact, you need to transform the line of a financial instrument into this smooth equitability going smoothly to the sky by means of regularities (patterns)))
Repeatable, self-similar and eternal, as you want, patterns are plenty in forex, only it is not possible to earn on them )))).
Or if it is possible, can you tell me how?
FAGOTT:
try real trading with tacticadverse - you'll see for yourself.
Have you tried it? I don't need it, I have my own tactics, MTS in the sense of I gave TacticsAdverse models as a close analogue of the patterns I wrote about. For clarity. I'm too lazy to draw it, so I gave the link, it's all there drawn a long time ago and even discussed in detail.
You can't trade "according to Adverse Tactics", you can trade MTS written based on Adverse Tactics models. Completely different things. And here, as many people as there are MTS, although the models are the same. And the result, of course, will be different.
There is a certain process behind these patterns (and behind the Neo patterns too, that's their similarity, though the processes are different). What's the difference between a pattern that can be traded profitably and just a "curve" on a chart was discussed in detail in the Patterns thread. The importance of understanding the process that leads to the formation of a pattern was also discussed there.
Although the process is the same for everyone and the patterns are the same, you can participate in the process in different ways, therefore there are many different systems that can be built.
There are people who build MTS specifically on the Tactics Adverse models, some successfully (they say). You don't have to believe them, but you can always check, yourself. I always do.
P.S.
What are you looking for anyway? And in what form should the information be, that would make you say: cool! That's what you need! In the branches of Neo, not only were the methods of finding patterns discussed, but a ready-made pattern was handed to you on a silver platter, just take it and trade. What else do you need? You want him to trade it for you and wire the money to your account? Calling Neo and the participants of the Branch on patterns - sectarians. That was strong. I appreciate the magnitude.
a ready-made pattern, working, handed to you on a silver platter, just take it and trade it.