Using artificial intelligence at MTS - page 6
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
it is called an expert system, and the proposed approach is the simplest of all the theory.
At least think about what is a "lying turkey" and what is a "properly working" turkey?
It is a set of formal attributes that allow attributing an indicator to a particular type.
This is why expert systems often use fuzzy logic, which is also a whole science.
> However, if you know something more profound and effective about it, please speak up.
Nah, I'm an amateur at it, but I can give you an idea (though again, all this can be found in the literature).
We should determine the clusters of inductors, i.e. group of indicators producing signals synchronously (synchronously in fuzzy logic terms) that will allow not only to determine if the indicator is "lying" or not, but also to determine the relations between indicators.
For example, suppose that at high values of X0 indicator X1 starts to blatantly lie, while X2 indicator tells the truth.
And it is vice versa when X0 is small.
When the expert system detects this, it will start to run X1 indicator only for small values of X0 and X2 for large values.
Now, imagine that X0 indicator is a trend indicator ;)
That is, the system will automatically detect that X2 works well in a trend and X1 in a flat.
(But according to your method - it would just give them twos and turn them the fuck off).
And if you add genetics to it, you will get a monster.
Only it seems that mql won't be able to do it, because firstly, it's limited by performance, and secondly, it's unreal to do without debugger.
In your Expert Advisor's code, where the perseptron function is calculated, AC value from zero bar is used. This means that during testing the Expert Advisor is looking into the future, because it uses the current value of AC, which has not really formed yet. And this casts doubt on the objectivity of testing and the results of forward tests on the remaining history.
Pyh, he does not look into the future. The testing mode is by bar opening prices, i.e. here it is not necessary for zero bar to be fully formed at all. Yes, testing would be really biased if one of the two remaining modes were chosen - since in the real world the signals in the zero bar would be constantly changing (perhaps there really would be a peek into the future here). It seems that the zero bar test can only be adequately performed in this test mode.
P.S. I agree with Yurixx's opinion. Rudeness should not be tolerated, although the expert should be acknowledged as very curious.
Sincerely Pooh.
P.S. I agree with Yurixx's opinion. Rudeness should not be tolerated, although the expert must be acknowledged as very curious.
I am not convinced by you. I understand very well that testing is by bar opening prices, BUT ! It opens a bar and we have to find (for this EA) the AC value at four points, including the AC value of the bar which has just opened. Where do we get AC if it is formed only at the closure of the bar?
It (the open price of the bar) will not change during the formation of the bar (may change High, Low and Close, but Open - do not, because the bar is already open).
I hope it is clear:)
So, as it turns out, Reshetov's Expert Advisor is not only an AI and a neural network, but also a software-implemented clairvoyant. :-)))