Machine learning in trading: theory, models, practice and algo-trading - page 1899
You are missing trading opportunities:
- Free trading apps
- Over 8,000 signals for copying
- Economic news for exploring financial markets
Registration
Log in
You agree to website policy and terms of use
If you do not have an account, please register
What does that mean? If everything is done through MO, you don't need any take profit or stop loss parameters. They should be generated automatically.
Different rows have different statistical characteristics
I modified the algo and the MO module a little bit. The result is something like this. On the right training (do not shout that it is on the right, it does not matter here).
Optimize on 5-minute ticks or opening prices (it's faster on openings).
Anything before 2010, the algorithm does not know. This is a good area to check what you have optimized.
Added 2 parameters:
This is a strategy of returning to the average at 1-4 hours.
Why can't you go against the trend? If we go against the trend, the commission will eat everything...
oooh, beautiful
Why do peopledeceive themselves?
It only seems beautiful, but in fact many years do not earn anything.
It is also necessary to show the statistics.
It only appears to be beauty, and in fact for many years it earns nothing.
Art requires sacrifice. Painting is not for petty mercantile people. ))
I modified the algo and the MO module a little bit. The result is something like this. On the right side is training (do not shout that it is on the right, it does not matter here).
Here I will support Maxim. I, too, first have 20% control section, then 100% test section and 100% training. Actually, from the essence of the optimizer itself, test and learning are one and the same section mirrored between two polynomials. However I admit that the checking section is at the very beginning because I need to make sure the model is adequate to the market at the current moment and thus I don't lose precious time of the NC operation immediately after optimization.
There is a feeling that pure seasonality without taking fundamental data into account does not work.
There is a feeling that pure seasonality without taking fundamental data into account does not work.
I agree, but the background indicator is complicated. Although you can start with the stock indices. It is not enough, but at least something.
We need indices for the states, and there are only static quarterly figures, but or the rates of the Central Bank. How to reduce it to at least a couple of parameters ... can't figure it out yet....Guys this is a complete mess, I almost pooped my pants. Started to update the system and during the installation came out the "screen of death", it's good that everything turned out okay.
Immediately a question to the inhabitants, who started building OI with that advisor that I threw out. Max???? If so, skip the files in the archive and then I have formed a couple of large holes in the past few days. I will be very grateful to....
I agree, but the background indicator is complicated. Although you can start with the accounting of stock indices. Not enough, but at least something.
We need indices for the states, and there are only static quarterly figures, but or the rates of the Central Bank. How to bring it together even into at least a couple of parameters ... can't figure it out yet....Well, we have the economic calendar api)
This is the problem to determine the difference between the months/quarterlies, where the seasonality works - is quite a task of the Ministry of Defense).