Australian (ASX) Stock Market Forum

Expert Advisors - Q & A etc.

Jump to new
Algorithmic optimisation --- the new frontier for Expert Advisor programming. :cautious:

I have found minorize–maximize (MM) algorithms are derived from expactation-maximise (EM) algorithms.

I think this may be the basic algorithm (EM) that programmers use to determine lot sizes BUT i`m not sure.
 
BentRod said:
Fapy...Are you going to answer Norm and W????:cautious:
Click to expand...

What do you care.... You are obviously... what is the word... trolling....


"2. All members must treat other members with respect at all times. This means insults, name calling and personal attacks on other members are strictly forbidden. Forum trolling - the intentional provoking of other forum members - is also forbidden. Please treat other members as you yourself would wish to be treated. Repeated violations of this rule will result in the suspension of your Aussie Stock Forums account."
 
fapturbo said:
What do you care.... You are obviously... what is the word... trolling....


"2. All members must treat other members with respect at all times. This means insults, name calling and personal attacks on other members are strictly forbidden. Forum trolling - the intentional provoking of other forum members - is also forbidden. Please treat other members as you yourself would wish to be treated. Repeated violations of this rule will result in the suspension of your Aussie Stock Forums account."
Click to expand...

nah i think the term your after is ..." sniffing bullsheeet" dunno what rule number that is but a few here have an uncanny nose for it ......... hey if he,s trolling as one would say .so be it ...but dont the truth set u free turboman?
 
Wysiwyg said:
Algorithmic optimisation --- the new frontier for Expert Advisor programming. :cautious:
Click to expand...


there is a really simple EA which can take $10k and turn it into $1.5 billion.


...on a backtest.

after full optimisation.

and increasing lot sizes.

u put it on a live account. itll kill it in 2 days :p: . everytime. i think the longest its lasted was just under 2 weeks.
 
BentRod said:
Just to add, I read on FF:



Was wondering if the same happened to you or still going by demo results?

(serious question!)
Click to expand...

I have activated my EA with Go Markets. I have traded it on live account before they declared it not to be used.

No I did not take the losses you refer to.

Oh and so far the EA has accumulated more pips than it has lost.

Believe it?

Go Markets as you are fully aware have closed it down so it cannot take trades.

Now I don't know what part of my posts are ...... "sniffing bull****"..... bull**** so if you think I am full of it then say so and give me a reason to believe you.

Just come out and ask serious questions without all the bull**** in the first place.

Like I've posted before I am not here to ramp the product. Just here to share my experiences. If you don't want to know about my experiences then don't waste your time reading the "bull****"

You obviously believe mechanical style Trading Systems which are fully automated are doomed to lose money in the long term.
 
Wysiwyg said:
Algorithmic optimisation --- the new frontier for Expert Advisor programming. :cautious:

I have found minorize–maximize (MM) algorithms are derived from expactation-maximise (EM) algorithms.

I think this may be the basic algorithm (EM) that programmers use to determine lot sizes BUT i`m not sure.
Click to expand...
maybe have a look at this Masters Thesis on the use of EM algorithms with Hidden Markov Models of Forex: http://liu.diva-portal.org/smash/record.jsf?pid=diva2:17431

From my experience these kind of algorithms, which can distinguish between local minima/maxima and global minima/maxima, are more important when deciding to take profits, rather than regarding lot sizes.
 
fapturbo said:
I have activated my EA with Go Markets. I have traded it on live account before they declared it not to be used.

No I did not take the losses you refer to.

Oh and so far the EA has accumulated more pips than it has lost.

Believe it?

Go Markets as you are fully aware have closed it down so it cannot take trades.

Now I don't know what part of my posts are ...... "sniffing bull****"..... bull**** so if you think I am full of it then say so and give me a reason to believe you.

Just come out and ask serious questions without all the bull**** in the first place.

Like I've posted before I am not here to ramp the product. Just here to share my experiences. If you don't want to know about my experiences then don't waste your time reading the "bull****"

You obviously believe mechanical style Trading Systems which are fully automated are doomed to lose money in the long term.
Click to expand...

i dont think it helps that your username is taken from a commercial EA product. it immediately sends up red flags to many.

can you run that FAP backtest with money management off so that lotsize is consistent?
 
pilbara said:
maybe have a look at this Masters Thesis on the use of EM algorithms with Hidden Markov Models of Forex: http://liu.diva-portal.org/smash/record.jsf?pid=diva2:17431

From my experience these kind of algorithms, which can distinguish between local minima/maxima and global minima/maxima, are more important when deciding to take profits, rather than regarding lot sizes.
Click to expand...

that is an impressive thesis. i have downloaded it and shall read it when i get some time.

congratulations on your hard work and research.
 
Stormin_Norman said:
that is an impressive thesis. i have downloaded it and shall read it when i get some time.

congratulations on your hard work and research.
Click to expand...


Yes I second that.Apart from the mathematical formula sections I found the thesis very interesting.Tells me that the big end of town have a better grasp on auto-trading than the (end of food chain) retail sector does.
 
You obviously believe mechanical style Trading Systems which are fully automated are doomed to lose money in the long term.
Click to expand...

Not the case, I just believe Fapturbo is doomed.

I am not here to ramp the product
Click to expand...

Your Nick is Fapturbo, your Avatar is a Gif from their site and you have "currently trading with ea......blah blah blah" written everywhere, sorry for thinking you are here to ramp ....LOL

BTW...how bout changing your sigs to:

"Not Currently using an Expert Advisor to Trade Forex"

and

"Not Trading FX with an EA"
 
BentRod said:
Not the case, I just believe Fapturbo is doomed.



Your Nick is Fapturbo, your Avatar is a Gif from their site and you have "currently trading with ea......blah blah blah" written everywhere, sorry for thinking you are here to ramp ....LOL

BTW...how bout changing your sigs to:

"Not Currently using an Expert Advisor to Trade Forex"

and

"Not Trading FX with an EA"
Click to expand...


I'm doing some extensive backtesting mate to really give this thing a work out.

At this stage I'm finding out that this EA is not as good as reported. However over the long term it does make money but there are periods of drawdown which could be a problem.

I'm being up front here and not trying to bull****.

Please see the backtested report here. This is achieved using 90% Modelling

http://www.sendspace.com/file/k18ufi

Using a high leverage setting during those periods of drawdown would really hurt an account.

Anyone considering using this EA needs to be fully aware of that.

55456263cy1.jpg
 
:D So what you are showing us is ... every dog has its day.Seems like knowing when to switch the thing on and off is worth examining.
:2twocents
 
I don't know what sort of optimizations are done on this EA.

Seems to be some curve fitting has been done for recent periods which has given this EA some sort of profitability.

Too much curve fitting can be a problem.

Like I've said before the only real test is to do forward testing.

Having backtested gives some idea on how it has worked. Does not mean it will continue to work. The market is ever changing.

I don't know how you determine if a system is broken. What sort of drawdown would you take before you decided to stop.
 
Top