Machine learning in trading: theory, models, practice and algo-trading - page 830
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
I agree, idle talk, no results.
I'm talking about the results of the model run, not the results from the real world.
What's the problem with that?
The first option is unambiguous, if you're talking about me, wiped his posts because he was ashamed, embarrassed, who are "they" I do not know
JB also cleaned up after my comment to his message
well wizard is paranoid, he does it all the time :)
Advice, then.
Once again.
1. Set R- 5 min.
2. Set rattle() - 5 min
3. taking my article as a guide and ready file for experiments - 30 min
4. We test the file, which is attached to the article.
After that you will ALWAYS be able to make perfectly concrete assertions and give answers to so many questions in this machine learning forum.
P.S.
If you have a desire, using the log kept by rattle() you can disassemble the finished code in R, which you generate with the mouse and maybe write something of your own.
In general, of course, laughing is laughing, but if the coryphaei meant the things I thought they meant, then they are reasonable and robust
Ah, that's it! The coryphaei have stolen the Grail!!! I am in sorrow... I sprinkle tears on the Russian land, the Orthodox ...
Ah, that's it! The koryphaei have stolen the Grail!!! I am in sorrow... I sprinkle my tears on the Russian, Orthodox land...
Well, it's not exactly the Holy Gra il, but we can already try to work with it...
Ah, that's it! The koryphaei have stolen the Grail!!! I am in sorrow... I sprinkle my tears on the Russian, Orthodox land...
AsMaxim Dmitrievsky used to say, these are all tests, and they're about nothing. As for myself, I can add that the tests, especially the MT optimization, are worthless.
Correct tests are very complicated thing, and, imho, only self-written ones are possible. I don't even want to talk about optimization.
Let's say it's not the Grail, but you can already try to work with it.
We must try it. And the trading signal, let's have it!
The coryphaei's song about their unrestrained love for money is completely un-Russian. The Russian market is not that bad and it is disgusting to read their opuses and whining.
No one is obliged to reveal their secrets, but signal on the barrel! Let people use it.
We must try it. And the trading signal, let's have it!
The coryphaei's song about their unbridled love of money is completely un-Russian. The Russian language is not the best way to read their opuses and whining.
No one is obliged to reveal their secrets, but signal on the barrel! Let people use it.
when I'll be at least 80% sure I will not disgrace myself - I'll do it :) while tests and modifications
As soon as I'll be at least 80% sure that I will not disgrace myself - I'll make a system :) until more tests and modifications
That's the deal, Maxim!
AsMaxim Dmitrievsky said, these are all tests, and they are about nothing. As Maxim Dmitrievsky said, tests, and even more so optimization in MT, are all for naught.
Correct tests are very complicated thing, and, imho, only self-written ones are possible. I am not even talking about optimization.
I think he was saying that tests (over a convincingly long period) without OOS and without signal confirmation are worthless