I don't think it the use of neuron but rather indicators are a combination of mathematical function.

So it is a generation of combination of mathematical functions that create  "indicators" for the use of strategy generation.

That way the strategy generation goes to more fundamental level.

All indicators are mathematical operation of the 4 basic data, namely O,H,L and C and possibly its previous value.

I would proposed the generated indicator be based on its ability to assist in generation of profitable strategy.

Once we have a few of this generated " indicators ", we can then use the normal strategy generation to generate hopefully more robust strategy.

I cannot go into much detail. Evolutionary algorithms were once part of FSB's generator and they turned to be inapplicable for the purpose. Popov "invented" the current generator after many trials and errors so you can be sure the options that come to mind first (or second), are pretty much tested and added, if they have any value to the process.

ALR system

Sometime back I post a wish for a loss recovery system in FSB Pro. I just come across this forum http://www.stevehopwoodforex.com/phpBB3/viewtopic.php?f=23&t=3682&start=180 that has an EA that is trying to do that. I have just try to test it out but am not sure of its effectiveness.

If it do help, then consider building it as an option into the strategy used for actual trading by FSB.

Portfolio

Export

My Strategy involves several strategies and I would like to be able to export a sheet which would include a worksheet for each strategy and also a combined sheet.

Reason being....... I would like to be able to check total exposure as each day passes and also to monitor the stagnation periods.

Total exposure....... I mean the amount at risk when all of the strategies are trading, perhaps I have to check to see that I am not over extended with multiple trades in play.

Stagnation..... I want to see if there are days when there are not enough trades and need to use a different combination of strategies.... determine which strategies are stagnant on each day and perhaps develop differernt strategies.

So........ would it be possible to add an export format that would accomodate this, or similar.

For my purpose, I would not need all of the fields in a journal sheet, but  the first few would be great.

Other people may have ideas for preferred output, perhaps we can gather ideas from all the users before any changes are undertaken.

Thanks for consideration

daveM

I want to ask if there is an option to make the trader ignore Saturday and Sunday data.
Also I want to ask how many bars is recommended for the MT4 when is
connected to the trader. Do these loaded bars affect a lot to the computers speed and memory?

Also I want to ask if there is an option for the trader to send pending orders with a predifined
difference between the entry signal and the entry order.

This is not implemented. You can make it avoiding trades on Sunday evening but it always uses all data provided by the trading terminal.

It depends from the indicators you are using. You can safely set 1000 or 2000 bars. If  you have only 2-3 strategies there are no problems. Eventually the program may start lagging if you are trading more than 8-10 connections.
You can solve that by starting several copies of FSB Pro and to distribute the load. It also depend on your machine. The more powerful is one, the better.

Some indicators have a vertical shift option. You can use them.
One simple example is the Price Move indicator. You set a price move distance from the Open price.

The program takes the time from the historical data instead of the local computer time.
You simply provide data files from the broker you use and the time will be correct.

There is no reason for them to be.
Why would you even think they are dependent on PC time?

Forex Strategy Builder was not created overnight. It was continually tested by thousands of users over more than half a decade.

You can trust the process.

Personaly I don't like strategies that are flat for a long time.
It could be added a strategy filter to FSB Pro that accepts for example only the strategies that
are profitable from day to day, from week to week, from fortnight to fortnight, from month to month etc accondingly to the users desire.
Also it could be added a filter that accepts the less daily weekly monthly etc profit accepted.
I am not talking about mean value but the lower bound of profit accepted.
For example I want my generated strategy to be profitable every any two months interval.
Also this time frame could be consecutive IanFeb MarApr MayJun JulAug
or overlapping like this JanFeb FebMar MarApr AprMay MayJun JunJul JulAug
Personaly I would prefare the overlapping option and I would prefare to have
profit at least $100 per interval.
An other idea would be acceptance criteria applied to these intervals like win/loss,
number of trades, consecutive losses, long short etc.
Also about my two month interval it could be applied not from month to month but
day to day week to week etc. I mean 1stJan-1stFeb 8thJan-8thFeb 15thJan-15thFeb etc.
This way FSB would generate consistent strategies with no long stops or pullbacks
and stable profitability minimum among a desired interval all along the time.
I must say that I am doing all this work my self with my old fashion mathematical tools
but it worths the effort.

Let us suppose that FSB generates a strategy from some historical data.
The strategy comes with some statistics based to these data (backtest).
win/loss ratio, average pips per transaction, dd in pips, stagnation
(I think these are the most important but somebody could add some more)
It would be helpful to save these statistics on the generated strategy file description
and for FST when trades this strategy live to have a page with a continous report
about divergences or convergences between live trades and backtest statistics.

This means that I believe that FST can be expanded and stay necessary for a long more.

Ehm.. this is already done with OOS (out of sample) testing. Basically OOS allows you to completely hide part of the newest data from the generator. So it operates with the first 80% of the data (example) and you can see how the strategy preforms on it's own on unknown to it data.

it is described here http://forexsb.com/wiki/fsbpro_guide/st … sample_oos