Machine learning in trading: theory, models, practice and algo-trading - page 1768

 
Mihail Marchukajtes:
I have a fully working technology of data preparation and it works on the data I use. Delta, volume, price. So I propose to join forces and expand the scope of the data I get from the market, thus improving the system that works well as it is. How about you?

Ready to participate/help in any way I can (AI/ML/DL/Qpile).

 

I have something similar in mind:

TC recognizes a sequence of similar patterns next to each other.

...and makes a forecast with respect to destroying the pattern. Probably at least two NS should work. SOM classify similar vectors (vectors from a floating "visible" window of some width (by time). We work, for example, with the most dominant class. Well, and MLP, can already predict fidelity/falsehood of the new pattern, based on statistics (for example) of such repetitions, taking into account all sorts of factors.

About cooperation, then, I don't mind. Rather, get tired of my inefficiency. My code formulation and debugging process stretches sometimes for a long time...

What Alexey Martyanov said long time ago, I think will be relevant for a long time to come... https://www.youtube.com/watch?v=0BwEeOyUa9w&t=734s

Философия трейдинга от My Trade
Философия трейдинга от My Trade
  • www.youtube.com
Призовой фонд $100.000 и 100 эксклюзивных Secret Box от сообщества SECTA на бесплатном турнире для трейдеров. Регистрация для участия в турнире http://bit.ly...
 
Sealdo Sergey:

I have something similar in mind:

The TC recognizes a sequence of similar patterns next to each other.

...and makes a prediction about the destruction of the pattern. Probably at least two NSs should work. SOM classifies similar vectors (vectors from a floating "visible" window of some width (in time)). We work, for example, with the most dominant class. Well, and MLP, can already predict fidelity/falsehood of the new pattern, based on statistics (for example) of such repetitions, taking into account all sorts of factors.

About cooperation, then, I don't mind. Rather, get tired of my inefficiency. The process of code formulation and debugging at me sometimes stretches for a long time...

Sorry, we seem to be from different planets on the MO with you, nothing personal, but your approach is flawed and unworkable. I'm sorry... All the patterns work 50/50 and it will be very difficult to understand them, the main thing is not to make mistakes...
 
And philosophy has nothing to do with it... not to put it on here, sorry....
 
Mihail Marchukajtes:
And philosophy has nothing to do here... I'm sorry, it cannot be attached here....

Welcome back))) Hope viruses don't bother you anymore)))) Good you)))) At least explain your approach, what is the difference, I understand about the lissajous figures, but I do not understand the essence of it))))

 
Valeriy Yastremskiy:

Welcome back))) Hope viruses don't bother you anymore)))) Good you)))) At least explain your approach how it differs, about the lissajous figures I understand, but I do not understand the essence further))))

Figures is a long-standing passion, which turned out to be unpromising. So I work in a different field, well, as for viruses, so it seems to have remained, but I now try not to publish posts with advertising. I have to remove them manually. All my advertising is only optimizer Reshetov Yuri with my modification and no more and then I do not sell it, and so ... just bragging....


By the way guys, you won't believe it, but I recorded a pilot sample note, I think it even went well, but I can not publish it yet, first a bunch of left-handedness, secondly Ubuntu cuntu not read the format goproshka, neo I'll solve it in the next few days. And if you do not mind, I'll publish the pilot, it's all sincerity :-)

 
dr.mr.mom:

Ready to participate/help in any way I can (AI/ML/DL/Qpile).

Get some Validol, good man.... )

 
Mihail Marchukajtes:

Figures is a long-standing passion, which turned out to be unpromising. So I work in a different field, well, as for viruses, so it seems to remain, only I now try not to publish posts with advertising. I have to remove them manually. All my advertising is only optimizer Reshetov Yuri with my modification and no more and then I do not sell it, and so ... just bragging....


I made a pilot note, it looks good, but I can't post it, first of all I have a lot of left signals, and second Ubuntu does not read Goproshka format, but I will decide about it in the next few days. If you don't mind, I'll publish the pilot, it's sincere :-)

I looked at Reshetov's profile askew, are you looking for a middle ground in volatility?

 
mytarmailS:

Stock up on Validol, good man.... )

Actually controversial statement, so I am kind and immensely loyal, I just do not have enough alternative data for my TC. And my approach feeds me now. I eat exclusively from exchange market, I will not say that straight lobsters, quite often doshik gets but the fact !!!! And what can you boast! (Dude, I do not mean you specifically, but in general, I ask people)

:-)

 
Valeriy Yastremskiy:

Looked at Reshetov's profile obliquely, are you looking for a middle ground in volatility?

No, I use his development of JPrediction which I updated to my own needs. Why?