Machine learning in trading: theory, models, practice and algo-trading - page 685

 
Mihail Marchukajtes:

Unfortunately I cannot agree with this statement. To the input of NS it is necessary to supply such values, which are the cause for the price, and what distribution they will have is absolutely unimportant. Because the found model, even if it has a non-Poisson distribution of inputs, will work in the future, because the nature of the input data will be the cause for the output.

And I agree with the statement that the more we transform the input data, the worse it becomes. The data should be taken as it is, without change. Unless minimum normalization by lag subtraction. Any other complication of input calculation leads to loss of notorious alpha and possibility of forecasting. IMHO of course. So... thoughts aloud.....

+100

Although, the window problem remains

 
Vitaly Muzichenko:

Time plays no role in trading, there's no need to get attached to it at all.

If there is no cyclic pattern, it is there, daily, weekly... It is very well seen in ZZ.

 
Renat Akhtyamov:


Once again.

For forecasting, it is extremely important, incredibly important, to know the laws of distribution of the predicted values.

You don't know them for prices, or for increments, or for the time between quotes. Moreover, you do not even try to bring them to one form or another. So how can you predict? These shameful tick archives have already plowed through a billion traders. Result = 0.

I've worked with it just a little bit and I'm in the black every week. I almost grabbed the Grail by the ears yesterday (it turned out to be my Schrodinger's cat...)

 
Vizard_:

Finally. It's easy to check. The model is trained without and with the time feed...

Can you expand on the methodology of the experiment a bit?

Because I haven't got it so far.

 
San Sanych Fomenko:

If there is no cycle, but there is, daily, weekly... You can see it very well in ZZ.

Cyclicality exists, but it cannot be predicted by time. The price can move an hour or a day, or even a month. It is possible to forecast the price passing but without time.

 
Renat Akhtyamov:

Are you sure?


Now I'm not sure. You and I are moving in the same direction. Now I'll just get the time and the same picture will be even better.

 
Alexander_K2:

Now I'm not so sure. You and I are moving in the same direction. Now I'll just figure it out in time and the same picture will be even better.

I have arrived, which I sincerely wish you to do as well.
 
Renat Akhtyamov:
I'm here, and I wish you the same.

HAHAHAHAHAHAHAH... For a long time now, by the looks of it.

 
Vitaly Muzichenko:

There is cyclicality, but it cannot be predicted by time. The price can go an hour, a day, or a month. It is possible to predict the passage of the price, but without time.

We are talking about a different representation of quotes, not in a linear time... Mathematically, it's as natural as going to the toilet.

And Alexander writes absolutely sensible things, which few people here understand because of the lack of education.

 
Maxim Dmitrievsky:

We are talking about a different representation of quotes, not in linear time... mathematically it is as natural as going to the toilet

And Alexander is absolutely sensible things that just few people here understand due to the lack of education.

Maybe they are sensible...

But we proved and justified everything at the university, and also did experiments on the laboratory. They didn't say, "Take it as truth, which does not require proof.