I have not. Time is not the issue when you want something to work right. I have Friday,Saturday,Sunday to work on Optimization.

When I try to backtest this I get a lot of ordersend errors, is it trying to hedge or something?

Any idea what the trading logic is behind the EA?

A brief look at the source code reveals that the EA is using neural network. In particular Perceptron
http://en.wikipedia.org/wiki/Perceptron

I test it with 1 week data. The result are 34 deals and 50% loss.
Is it possible to train with each .set file separetly and then combine the results in one .set file, or the right way is to train 1.set load the result in 2.set then load the result from 1.set and 2.set in 3.set ..... etc?
I hope to understand me (my English is poor)

I posted the .sets but you don't have to do them seperately.

I numbered them so you know in which steps you have to Optimize so you start off by optimizing the 1.set then uncheck those and optmize the 2nd portion of the EA displayed in 2.set and so on.

Yes same me I would like .set file too.

__________________
http://ezinearticles.com/?Learn-The-...tem&id=2533201 | Learn Forex

Hi all,

Where i can get M5 period data from 2008 10 01 ?

After running it for a week I was very impressed with the results.
Interestingly enough When I tried to back test on the same last week data I got different results. They do resemble actual trades, yet different and not that great as live trades. It tells me that there is something wrong with back testing engine within MT4.

I talked to one experienced programmer, he told me MT4 has some serious flaws at its kernel and therefore all programming and optimization is more an art then science. So to design a real good EA in MT4 one has to know all its flaws in order to compensate, same applies for optimization.
MQ is going to come up with MT5 that hopefully will address that.

I also read original posts for this EA in Russian with all author remarks, they test it on other currencies, but so far EUR/USD is the most reliable. Author also mentioned, it's a beta test version and not yet suitable for live trades as error order checking isn't yet implemented in the code. You can still trade it with the account copier . Trade copier will copy trades from one account to another and will check for errors. Errors may occur in fast market, dew to re quotes or if you trade other systems on same account. I have trade copier and it works fine with that

I try to determine if this system worth running, because there are so many systems for EUR/USD with less optimization hassle. Normally all you need is 3-4 weeks of historical data. Normally 3:1 history to forecast is the minimum data req in statistics.
*****************

George, keep up your great work

Hi verybody.
I`m trying to test my trade system in historical data, but I really don't know how to do it. Could you give-me some help? Can I export graphically any data on the history center?

Thanks

Hi all,

Backtesting and optimization get you started in the right direction. However, given changing market dynamics over time, todays high performance EA may fail to perform at some point in the future. We all know this.

My preference is to backtest and optimize over a 1 year period, then test this on previous random years up to 8-9 years back to get a feel for overall performance. If I like the results, I will then zero-in to more recent trading history.

For example, I use MAs on the Daily USD/JPY. I take the slower MA and multiply this by 6 (7-day MA = a 6-week optimization period, for example). Then, at the end of each trading week I re-optimize by dropping the first week, and adding the week just completed. It is effectively a moving average applied to optimization.

Let me know your thoughts!