A quick and free library for MT4, much to the delight of neuralnetworkers - page 36
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
Here's how strange it is to me about optimisation, "Let's say 1 Jan. 2010 - November 1, 2010 - 10-month optimization period; November 2, 2010 - December 2. 2010 - 1 month - 10 % forward (OOS). % - you are the forward test". The instructions there also describe it as follows. Suppose we opt for one year from 01.01.2008 to 01.01.2009, then we will choose the most suitable parameters, then we will run forward for the period from 01.01.2009 to 01.04.2009 (let's say 1 April 2009), we will forward 3 months (25%) and find parameters that show good results on the forward, i.e. after the optimization period it will not pour with these parameters but here we have a question: where is the guarantee that it will not pour from the 4th month? Is not better to fill up at the most spoiled by today's date (April 01, 2009) and tomorrow put on the auction?
https://championship.mql5.com/2010/ru - eh, I'm rooting for the bunting:)))
Here's how strange it is to me about optimisation, "Let's say 1 Jan. 2010 - November 1, 2010 - 10-month optimization period; November 2, 2010 - December 2. 2010 - 1 month - 10 % forward (OOS). % - you are the forward test". The instructions there also describe it as follows. Suppose we opt for one year from 01.01.2008 to 01.01.2009, then we will choose the most suitable parameters, then we will run forward for the period from 01.01.2009 to 01.04.2009 (let's say 1 April 2009), we will forward 3 months (25%) and find parameters that show good results on forward, i.e. after the optimization period it does not pour with these parameters but here we have a question: where is the guarantee that it will not pour from the 4th month? Isn't it better to fill up to the barest limits on today's date (April 01, 2009) and put it up for auction from tomorrow?
That's how you do it... And trading period = forward... Read here - https://www.mql5.com/ru/forum/107064 and Robert Pardo "Designing, testing and optimising trading systems for
For more details see "The Market Trader" - very informative and instructive. I recommend it.
That's how it's done... and the trading period = forward... Read here - https://www.mql5.com/ru/forum/107064 and Robert Pardo "Developing, testing and optimizing trading systems for
The "Stock Trader" is very informative and enlightening. I recommend it.
I.e. it's better to optimize a year, then forward three months, and then three months later optimize again with a shift by three months? I've read this article, I'll find Robert Pardo on the web and read it:)
I just tested my EA (not FANN) and did the following: I took a period from 01.01.2010 and reopted it till 27th of November. On 29th of November I put it for testing from Monday and it is working with profit (for now))), but a week later I reopted it again from 01.01.01.2010 to 04.12.2010 years, ie added stupidly week and again prooptil and again hung the best results in the test, too, so far in the plus, now optimize again, again added a week, ouchu from 01.01.2010 year to date.....to such a method .....
I.e., is it more correct to optimize for one year, then forward for three months, and then, after three months again optimizing with a shift by three months? I've read that article, I'll find Robert Padreau's one on the web and read it :)
I.e. it is more correct to do optimization for a year, then forward three months; then again optimization for the period - for the same year + 3 months (forward) - then real with parallel demo (to trace
discrepancies and their possible reasons) - (3 months)...
Provided there are enough trades, min. 200-300 or more...
I.e. it is more correct to do optimization for a year, then forward three months; then again optimization for the period - for the same year + 3 months (forward) - then real with parallel demo (to trace
discrepancies and their possible reasons) - (3 months)...
Subject to a sufficient number of trades min. 200-300 or more...
I get it, i.e. after three months of real trading, we don't discard the "old" three months from the tail, but just add the forward and optimum again. I got it all, but there is a question here: where's the guarantee that it won't cast after forward? And my robot has traded an average of 2000 a year.... And about the book, Trader Joe's, I'm just crying :))
I get it, i.e. after three months of real trading we do not discard the "old " three monthsfrom the tail, but simply add the forward and optimum again. I got it all, but there is a question here: where's the guarantee that it won't cast after forward? And my robot has traded an average of 2000 a year.... And about the book, Trader Joe's, I'm just crying :))
I get it, i.e. after three months of real trading we do not discard the "old" three months from the tail, but simply add the forward and optimum again. I got it all, but there is a question here: where's the guarantee that it won't cast after forward? And my robot has traded an average of 2000 a year.... And about the book, Trader Joe's, I'm just crying :))
Yeah, there are so many grades in the book, it's crazy))) It's a real bummer there:)))