Machine learning in trading: theory, models, practice and algo-trading - page 3244

New comment
 
mytarmailS #:

Why aren't the strings supported????

https://www.mql5.com/ru/docs/onnx/onnx_types_autoconversion

Don't ask questions like, "Why do you need strings?"

That's why we want to launch the championship to bring everything to mass use.

Of course, we will expand the functionality.
 

I asked katbust developers if it is possible to change the second output to tensor instead of zipmap in a simple way - silence. There is no disabling button via onnxmltools either, other boosts have it. You can use sklearn2onnx, but it's ugly - a lot of code.

Although everyone realises that this method is inefficient in terms of speed, but they still do it for some reason.

 
fxsaber #:

The advisor has nothing to do with MOE. This is a technical test of how the proposed template works.

Why to go to the MO thread and report that MO is not working - I don't know.

What's the point of the tambourine dance?

 
Renat Akhtyamov #:

What's the tambourine dance for?


Who dances the tambourine?
 
It's getting really, really bad out there
 

Has anyone tried networks like PINNs (Physics-informed neural networks)?

It looks like you can drive formulas into their layers by constraining the gradient vector, which seems interesting.

Whether ONNX supports them is also a question....

 
Of the neural networks, the best one for BP is TimeGpt, so far. Which kind of hints, based on experience with chatgpt, that miracles don't happen :)
 
By means of bards and gpts I tried to speed up my code in python, which is not available in the internets, that is, without peeking. There are 20-30 lines. None of them worked, moreover, none of the proposed codes works at all. No matter how hard I try, there is no result. Which kinda hints too that these llm's don't know how to extrapolate.
They sort of understand that it is necessary to do vectorisation and connect numba, but the code is meaningless.
 
Maxim Dmitrievsky chatgpt, that miracles don't happen :)

is there a comparison test to say that?

or the usual...

 
mytarmailS #:

is there a comparison test to say that?

or the usual.

There isn't for you.
1...323732383239324032413242324332443245324632473248324932503251...3591
New comment