Better NN EA - page 6

 

Also there is a big number as order comment, I think that those numbers mean something for the Author, probably there is the number of the rule that triggered the trade.

Most NN software out there only output, exe, DLL Vb, C++, xls code, but as far as I know Metaquotes don't allow DLL in the championship, it mean that the Author had to write a library of rules in MQL code.

Probably he spend a lot of time to do it.

I bet that this EA have a library with over one thousand of rules in the NN logic, or at least that's what the Author want us to think.

And last, this EA appear to work on M1 timeframe, Metaquotes use unfiltered datafeed, but most broker datafeed is filtered, it mean that the Indicators who feed the NN logic can output a totally different signal, making the EA unprofitable on live brokers, try to run a RSI on M1 at metaquotes and run the same RSI at M1 on IBFX live, you will see how different is the output.

There are many pro and cons about this way of trade, in live accounts it can work, or maybe not.

Files:
betterea_1.gif  53 kb
 
oilfxpro:
Hi

This guy is really showing off his EA

Amazing!

Participant Better - Automated Trading Championship 2007

When does the competition end?

Oilfxpro

it ends on December..

his EA is really great, i think it is the best EA that join the competition since how many years.. if that really runs on live.. he can be a millionaire..

 
project1972:
The NN logic is not linear, Better EA sometimes trade Counter-Trend and Sometimes Trend-Following.

You can't back-engineering a non linear logic, most of the time not even the author know what are rules learned by a NN logic, it's a black-box for everyone including his own creator.

You can check the past trades, but you don't have a clue what rules will trigger the next.

from Better interview

His EA consisted of Three subsystem, No one of them "sees" what other subsystems are doing.

Perhaps subsystem means trading system or some kinds of methods to open trade... the NN calculates what system should open trade base on the highest probabilistic to win..

That's why sometimes Better EA trade Counter trend and trend following, I think

Pj..

 

------------

 
primajaya:
from Better interview

His EA consisted of Three subsystem, No one of them "sees" what other subsystems are doing.

Perhaps subsystem means trading system or some kinds of methods to open trade... the NN calculates what system should open trade base on the highest probabilistic to win..

That's why sometimes Better EA trade Counter trend and trend following, I think

Pj..

Means the NN is the brain of decision, instead of brain for trading descretion? Put it this way, NN detects a trending condition, according to the current ATR make sense that it is possible to ride the trend with low risk, so it trade with MA, ADX and shit. Whilst if the NN detects a divergence, it give up the trend follower and open a anti trend? Meanwhile, it could still scalping at all time(besides of spread widening time).

This makes the NN had a decision to open trade on Long Term, Mid Term and Scalping. All of them can open position of 5.00 lot at maximum simultaneously. It could anytime hedged itself because they all have different TradeComment / MagicNumber to seperate them.

Am I dumb or what? Is that what it suppose to be? How are we going to move forward if thats the case? I have no experience with NN at all. But the performance really shocking!

Regards

David

 
primajaya:
from Better interview

His EA consisted of Three subsystem, No one of them "sees" what other subsystems are doing.

Perhaps subsystem means trading system or some kinds of methods to open trade... the NN calculates what system should open trade base on the highest probabilistic to win..

That's why sometimes Better EA trade Counter trend and trend following, I think

Pj..

Three sysytems put together trend ,counter trend and triple screen.

As of next month I am running a forward demo test on the three sub-systems I own and I will get the same results as Better

These systems are already producing around 400 pips a month ,which Better's EA is producing

These systems are

oilfxpro countertrend

oilfxpro rsioma

oilfxpro supertrender

My biggest problem on this forum is getting a reliable coder to do a professional job.I am dissapointed with half a dozen of the best coders on this forum and I have tried the best of this forum.

OILFXPRO

 
project1972:
The NN logic is not linear, Better EA sometimes trade Counter-Trend and Sometimes Trend-Following.

You can't back-engineering a non linear logic, most of the time not even the author know what are rules learned by a NN logic, it's a black-box for everyone including his own creator.

You can check the past trades, but you don't have a clue what rules will trigger the next.

Project

You can backtest 3 different systems and puthem together

OILFXPRO

 
oilfxpro:
Project

You can backtest 3 different systems and puthem together

OILFXPRO

I think that you don't understand how a NN logic work.

In a linear system like yours, you create the rules.

With a NN logic, the NN engine find the rules, it can create a library of thousand of rules.

Not any Linear System can compete against a properly designed and trained NN logic, but to get to this point, WE are years behind, probably the Author of better EA have many years working on that system.

This is not something to hire a coder and wait a week to code it, I bet that there are many thousand hours behind that development.

I am also a newbie on NN logic, probable someone can explain it better than me

 

I find this interesting - how can we operate in an environment we we are paid to predict the future and yet convince ourselves that we do not predict future price?

Is it not true in fact that when I open my next buy, I do so because I believe price may go up? (In fact, I predict that price may go up.)

Just like the gun example - if I am correct, I get paid and can live to fight another day, if not, I am broke/dead (and out the game really)

As such, in my opinion, trading is a game of forecasting, predicting where price may go next.

Some of us may use fundamentals, others may use TA and still others may use NN's - in the end, we all try predict where price is likely to go next.

Well - my ideas for what it's worth.

 

Command line tester used by Better:

From past comments, it appears that some folks in here are interested in the development approach used by Better. They might find this sample program and article draft I wrote helpful.

In this article, I provide and briefly describe a sample command-line C program that includes: parsing of MT4-exported CSV files, collection of statistical information, a simple trading system based on the collected statistics, a trading simulator, and trivial parameter optimization loops around both the statistics collection and the trading simulation phases.

Profile Better - Automated Trading Championship 2007

Strategy Testing & Optimization in Standalone C Programs

Files:
fx2.zip  4 kb