Machine learning in trading: theory, models, practice and algo-trading - page 2047
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 answer is wrong, think again)
Well, formalize it first)) Translate it into machine code, can you do it? I could not, although it seems simple... I'll do through pattern recognition... MANually mark it up.
Well, at least try to think how it can be programmed, try hard)) And then you'll understand my main idea! That you can't teach such patterns with conventional approaches...
Neither normalization in its usual form nor a sliding window works here ... anyway, I wrote out all 5 points
Why isn't it right?
the answer is not correct, think again)
Well you formalize it first)) Translate it into machine code, can you do it? I could not, although it seems simple... I'll do it through pattern recognition... MANually mark it up.
Well, at least try to think how it can be programmed, try hard)) And then you'll understand my main idea! That you can't teach such patterns with conventional approaches...
Neither normalization in its usual form nor a sliding window works here ... in short, I've written out all 5 points
If I wanted to make one pattern, I could do it without the grid.
Sfigal is impossible. Put Mashki in the ns and teach how to trade only under these conditions
(This is where you don't need ns)))
Sfigal is impossible. Put Mashki in the ns and teach how to trade only under these conditions
If we know the pattern and have formalized it correctly?
Do we know the pattern? Do we formalize it?
There may be 5 000 of such patterns and we don't know any of them and haven't formalized them, but the net wants to find them all by itself))
Remember recently I gave you dataset with pattern/sequence ??? It's the same problem by the way
and no trasformer can do it....
Why isn't it right?
5 points wrote
Well that's if we know the pattern and we formalized it right ???
Do we know? Have we formalized it?
But there could be 5k such patterns and we don't know any of them and we haven't formalized them, but we want the net to find them all by itself ))
Remember recently I gave you dataset with pattern/sequence ??? It's the same problem by the way
and no trasformer is going to help here....
I don't want to look anything up, for example, because I'm stupid and lazy. I want us to find something.
I'm the same and I have the same desires...
I just wanted to explain by the example of that pattern that the net will never find anything for objective reasons...
It does not take into account the specifics of market data, datasets are done by the type of Fisher's irises
I'm the same way, and my desires are the same...
I just wanted to explain by the example of that pattern that the network will never find anything for objective reasons...
The specifics of market data are not taken into account, datasets are done by Fisher's type of irises
Rnn or Transformer should pull any meaningful context out of the VR, if it is really meaningful. I certainly can't suggest anything cooler in this delicate situation. Allah give me the strength to figure this out to the end and I'll show you the mother lode.
Rnn or Transformer should pull any meaningful context out of the VR, if it is really meaningful. I certainly can't suggest anything cooler in this delicate situation. Allah give me the strength to figure this out and I'll show you the mother of kuzma.
Okay, uh... I'll keep my fingers crossed.
But most importantly, do you understand why the network can't find the pattern I described? )