Machine learning in trading: theory, models, practice and algo-trading - page 2531
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
Nice girl from daisyot banned)
I am proceeding from the same thing as Drimmer - I need some way to isolate the movements. Then, by analogy with the method I used in my article about gaps , I study statistics of returns to the starting point of selected moves.
I like this approach too, but instead of gaps I can search for any kind of patterns.
I like this approach too, but instead of gaps you can look for any patterns.
I agree, in my article I have already considered the intersession gaps in addition to the usual ones. You can take the tops of zigzags, some levels and so on.
She talks in foul language)
She's already been unbanned - she only talked for 24 hours)
Is there any sense in comparing the results of training on a sample of, say, 10 thousand examples with 1000 predictors and with a sample where the predictors for training are random binary values? Will the training results be comparable?
With the quotes and with the target at TP=SL it was also 50/50%.
There was a variant with 47.5% error, it looked cool, but when I connected it to the MT tester it turned out to be a fall instead of growth. It turned out that I did not take into account the commission, it ate these 2% benefits.
Here's thinking about how to account for the commission...
I wanted to add 4 pts to the spread. But this is not right. Sometimes TP and SL will be triggered by an overestimated Ask, not on that bar, where it should be in the tester, because of this the order of subsequent trades may change.
But the tester uses the minimum spread on the bar, it will also differ from the reality.
I have not figured out the best way yet.
There is a proposal to the MT developers.
In the quotes of bars to keep the spread not the minimum per bar, but as the difference of the maximum prices (High Ask - High Bid). In this way, we can calculate the maximum value of Ask.
The minimum spread per bar does not allow us to calculate anything.
And in this way we will obtain the upper price of the High Ask.
We will know exactly what the TP or SL could trigger on this bar.
At minimum spread this stop may be missed and testing will differ from real trading and from testing based on real ticks.
Thus, bar testing will be close to testing on real ticks.
It seems to me it would be interesting for all those who use the tester (both with MO and with conventional EAs).
Support us if this is a good suggestion. Maybe the developers will do it.
There is a proposal to the MT developers.
In the quotes of bars to keep the spread not the minimum per bar, but as the difference of the maximum prices (High Ask - High Bid). In this way, we can calculate the maximum value of Ask.
The minimum spread per bar does not allow us to calculate anything.
And in this way we will obtain the upper price of the High Ask.
We will know exactly what the TP or SL could trigger on this bar.
At minimum spread this stop may be missed and testing will differ from real trading and from testing based on real ticks.
Thus, bar testing will be close to testing on real ticks.
I think it would be interesting to all who use the tester.
Support this idea if it is good. Perhaps the developers will do it.
The real spread in minutes is of course a good idea, but they will hardly do so, because the beautiful advertising image of a small spread on Forex may be spoiled.
At most, they will offer to independently create a custom symbol with necessary properties on the basis of real ticks.
The "reference" quotes are from yahoo finance, because they take them directly from the interbank.
And the brokerage companies have mixed quotations - they mix and distort them.