Machine learning in trading: theory, models, practice and algo-trading - page 1083
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
hilarious...
the same
We know, we've heard:)
it?
https://www.forexfactory.com/showthread.php?t=278340
Put it in your TS and see. If the grid, you have to prepare it properly, otherwise it will blow the roof (levels).
here's cooking normally, it's not clear for me yet....
The usual extremums do not work, various complex combinations work along the way, but the grid will not find them. I need a separate algorithm that will gather the combinations and check them for validity and then feed them into the net... but there are many questions about the algorithm, I don't know yet how to realize it even at the block diagram level.
The trees and so on in the forest.
Well, yes, but this is the next step, and I'm stuck on the very first one - transformation of information, so to look for combinations you must first create them, those divide the data into certain pieces of clusters that can already be connected, added, thrown out in general combine... but how to do this, what to count as pieces I don't know...
Here's the usual simple pattern, but the two-turn, no network will not find it now draw ...
1) Wait for the zigzag movement
2) it's a multiway process, the price may move anywhere, the main thing is to break through 261.8
3) Only after that it is possible to expect a rebound from the 100% line, i.e. condition (2) may or may not occur at all
The rebound may not take place, but it's more likely caused by the fact that something is not taken into account when it (something) happens.
1) You see, I wrote a zigzag movement in human language, everything is simple, but what is the movement - a wave in the form of a code? How can it be programmed if the market is fractal? And you need to program it, and even bring it all to the same parameters to then share and do these combinations to find something...
2) It is not working to put it on the network, you understand.... And the network will not find multi-way situations and the market is only such ... multi-moves - when one situation grows out of another and so can last many times
Well 1 - you can probably zigzag, if the distance to 2 is enough to
to avoid redrawing. Look for articles - I think someone highlighted paternas zzz, etc..
There's no need for MO rattles here...
I've read about it and tried zigzags.
The question is not about how to describe the given pattern, there are hundreds of them, the question is how to make the net to find such class patterns, with motor variants, with unclear target stretching in time and so on...
the first thing is how to convert the data, so I'm thinking
Yes, I read it, and tried the zigzag, fractality must be overcome somehow.
And the question is not how to describe this particular pattern, there are hundreds more of them, the question is how to make the network find patterns of this class, with mohodovy variants, with incomprehensible target stretching in time, etc...
the first thing is how to convert the data, so I'm thinking
there are methods for estimating fractality, the most popular is box-counting dimension
time can be excluded with Renko graphs
there are methods for estimating fractality, the most popular being box-counting dimension
time can be excluded with Renko graphs
Renco is the same as a zigzag, you also need to enter the range parameter and this is not good
How do I see the fractality estimation, how do I use it?
Renco is the same as the zigzag, you also need to enter the range parameter, and this is not good
And about the fractality estimation, how to use it?
Renco is not a PZ, Renco will hide all the fine movements inside the bar - it works by breaking through the Renco brick by the value of the Renco brick, it lags, but it will not have a time component and there will be no noise components
You yourself started about fractality, I gave the method I read a week ago on hobrehabre
What does fractality represent? Nobody knows, I think that this estimation is similar to entropy, and all that can be found in fractality estimation or entropy is that there was a change in the market, but you can see it with the naked eye)))
I'm just studying the box-counting dimension, I haven't written it in code yet, I'll write it, maybe something will show up... but i doubt it
there are methods for estimating fractality, the most popular being box-counting dimension
time can be excluded with Renko graphs
Igor, didn't you show a distribution with ears? Lost it.
Let me see it again and tell me from what and how it was obtained.