Recognising images ( rhetorical theme ) - page 6

New comment
 
Reshetov:

The ToR must be specific. I.e. it is up to the customer to specify what a double top is. At the very least give us an example of the double top

and all other shapes that are not double tops. A concrete example could be a section of historical data on some instrument

The TOR should be very specific, i.e. from a bar with time so-and-so to a bar with time so-and-so we have such-and-such a figure.


In this case we make a block that distinguishes a double top from any other pattern. And then we form a trading signal according to the recognition results.

Simply put, we need to separate flies from cutlets, namely, to make two blocks: pattern recognition, i.e. classification of patterns, and trading signals forming according to the results of this recognition.

Neural networks are quite suitable for classification tasks.


YES ! that's right ! 100% ...

If you are writing on request, rather than free creativity with the customer

 
recognising is not a problem, if only one knew what to recognise - that's worth a lot...!
 
denis_orlov: It's not a problem to recognise it, if only one knew what to recognise - that's worth a lot...!
What's the tester for? We need to test the "adequacy" of the patterns. For me the tester is not a "mechanism for optimization", but "research equipment". I'll tell you a secret: most variants of "variant space" will leak :)))
 
SergNF:


Has your "Expert Advisor" always recognised the same "geometry" that you "trade" with your eyes?

I didn't.

My EA recognised patterns quite accurately. And much more efficiently than with my eyes, because eyes sometimes see what they want to see. When I ran it carefully through the history, the statistics of predicting the direction was close to 50%. And profits, if there were any, were intermittent, i.e. 2-3 months of successful trades were followed by 2-3 months of complete loss.

I don't want to argue with anyone. If you have a successful EA trading by patterns, you should post the statements. And then we will have a substantive conversation.

 
YuraZ:

No, I didn't... that's what the customers want.

--

it's hard enough to recognize an image... although there is the experience of mankind in this direction

If anything, send it to me, I recognize not expensive:), ate on this case a dog (and a cat).

 
Richie:
And what's the tester for? We need to test the "adequacy" of the patterns. For me, a tester is not a "mechanism for optimization", but "research equipment". I'll tell you a secret: most variants of "variant space" will leak :)))

I already know what the tester is for.

What I need to know is what to recognise...))

To recognize is a technical question, but "what exactly" is a universal one...

 

quote. who knows where it's from, don't say anything. i just liked it. :)

.... мужыки ВЫ такие умные ... можно с вами постоять на ветке ...киньте ссылку чтоб врубиться об чём речь ...

Also, off-topic, but also funny, sorry if anything.

... as in the early j.Reshetov's councillors....

PS these are real quotes. taken out of context. "do we not sometimes take out of context what should not be taken out, in order to understand the essence of the whole?" so I
 

Do you need to recognise patterns? Or images?

In the wonderful tale of Mulder and Scully there is an epigraph, verbatim,

"The truth is out there."

I subscribe. IMHO.

 
gpwr:

...the eyes sometimes see what they want to see...

Naturally, we are talking about the analysis of completed (manual) trades from the real, rather than hypothetical "I should have entered" trades.

I do not want to argue with anybody.

I support you, because :)

If we are writing on demand and not free creativity with the client

:|

 
When discussing this subject, you do not take into account the huge number of pitfalls.
12345678910111213...15
New comment