Machine learning in trading: theory, models, practice and algo-trading - page 986
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
Well, here we come to the Wiener process, or Wiener-like process. What's there to predict? The best prediction is the value itself.
But we can predict the normal distribution (or near-normal distribution). Not in the literal sense, of course, like a trace of a candle.
The Expert Advisor above. Prove it.
Counselor above. Prove it.
I am not interested in mrket, even free, and do not plan to.) Neither do I make models in MT, and I do not use a tester.
Can I use my model? Then look - a lot of very small deals. The trades are not random, but made on the assumption that the market is random. Sounds bad of course.))
Test of 3 months.
On X - trade number, on Y - profit in pps. chart.
I am not interested in mrket, even free, and do not plan to.) I don't make models in MT either.
Can I use my model? Then look - a lot of very small deals. The deals are not random, but made on the assumption that the market is random. Sounds bad of course.))
Test of 3 months.
By x - trade number, by y - profit in pps chart.
Everything is fine except for one important thing: where is the proof that it will be beautiful in the future?
Everything is wonderful except for a trifle: where is the proof that it will be just as beautiful in the future?
This is the future. On the model, of course. And there is no proof for the future, nor can there be. How are you going to get it?
I've been doing it for more than a year, and I think that this model will suffice for six months or a year. As O. Bender used to say - I don't need an eternal primus needle. I do not want to live forever.
The model, by the way, has not yet been implemented.
This is the future. On the model, of course. And there is no evidence for the future, nor can there be. How are you going to get it this way?
You're wrong, people even get nobility for such proofs.
On this subject I wrote many times
You're wrong, people even get nobility for such evidence.
I have written many times on this subject.
I don't pretend). But, if the model is tested correctly, as a rule in the real will be approximately similar results.
I don't pretend). But, if the model is properly tested, as a rule on the real will be about the same results.
Good luck.
For me, the result is a proof of future behavior of the Expert Advisor and not the profit of the Expert Advisor.
Good luck.
For me, the result will be proof of future EA behavior, not profits from the EA.
Boring, SanSanych. When the TS is implemented and works, there is nothing left to discuss - there is no subject. The development process is much more interesting.
What is there to prove? If the approach is acceptable to you, try it. If it is not acceptable to you - then please break with what you have. We exchanged opinions and dispersed, the forum, however.
For the appetizer.
A run of the same Expert Advisor with the same settings as above, but with a longer time interval.
This is the whole value of all these nice pictures.
The picture should prove the idea whose meaning is ONLY about the future behavior of the Expert Advisor.
The picture should prove an idea, the meaning of which is ONLY about the future behavior of the advisor.
In fact, that's what everyone does. A setup on one interval, a test on another. If the test is correct, the future behavior is almost guaranteed.
This is why I use my tester rather than MT's one - it seems to contain too many grails. In your own tester, at least reliably know - what and how he does. And information from the test can get a lot more and any, and easier.