Machine learning in trading: theory, models, practice and algo-trading - page 3004

 
mytarmailS #:
1)this is different

2)The type of MO may not be important, but the way information is presented is. As an example CNN's boom compared to past AMOs.

3) these are my words)

the words may be similar, but the meaning is different.

the very approach of training and subsequent prediction has a funny limitation, which in principle does not allow to do it.

Whoever guesses or googles why - will change his view on learning.

Naturally, we are talking about complex series, not a set of sine waves.

Absence of a signal in the data (hello Tsosniki) as an indicator of impossibility of such forecasting and vice versa, impossibility of forecasting in the classical way - absence of a signal.

But not the impossibility of forecasting in principle.

 
Maxim Dmitrievsky #:
What is this conversation about? Some vague statements.

Ps. A series of any complexity can be described by a set of sinusoids, just for reference))))
 
mytarmailS #:
What is this conversation about? Some vague statements.

Ps. A series of any complexity can be described by a set of sinusoids, just for reference))))
Because it's a puzzle, you have to ask neuronka to generate a picture.

You can only decompose it, but you can't predict it.
 
Maxim Dmitrievsky #:
Because it's a puzzle, you have to ask the neuron to generate a picture.

You can decompose it, but you can't predict it.
So what's the point?
 
mytarmailS #:
So what's the point?
That the m.o.d. Is being misapplied
 
Maxim Dmitrievsky #:
That the MoD is being misapplied
Well, it's clear from the result, is there a solution, or just to talk?
 
mytarmailS #:
Well, is it clear from the result, is there a solution, or just to talk?
The solution comes by itself, when there is understanding. Whoever understands why it doesn't work will come up with one.
 
Maxim Dmitrievsky #:
The solution comes by itself when there is understanding. Whoever understands why it doesn't work will come up with a solution.
I see.
So just talk.
 
mytarmailS #:
Oh, I see.
So just to talk.
It was a call to think before doing nonsense like transforming the original series into graphs, quanta, levels, sine waves or whatever else you can think of.
 
Maxim Dmitrievsky #:
It was a call to think before doing nonsense like transforming the initial series into graphs, quanta, levels, sinusoids or whatever else you have invented.
That the problem is not correctly set, I have been saying it here for a long time, the last time 10-20 pages ago, so there is nothing to teach me, I understand everything perfectly well myself...

And to transform the task, you have to transform the data...

So while you're telling me to think about it, I've already thought about it and I'm considering my options.