Better NN EA development - page 80

New comment
 

Big bugs!

barnix:
5 bar indicator

5_bar.mq4 (2.6 KB, 276)

5_bar_.mq4 (2.6 KB, 234)

I don't know how to can attached such scripts ?!

code:

for(int i=1 ;i<=NrOfBars;i++)

{

shigh=High;

....

it's mean that indicator takes data from future !

therefore signals good looks, but only looks.

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

But I see now that this indicator serves to train only, so then it's ok

 

I have started a C++ project: Better nn EA clone on google.

I want to make it free:

You can se the project at link:

Home (Better EA Clone)

You can find all C++ examples for designing a C++ command line tester.

For Better EA we must create a command line tester first in C++.

All project's it is in codeblocks.

On windows you must install Cygvin with boost libs.

 
OnTheRoad:
hi to all ,

these are good resources for NN library in C and C++:

1-FANN library by Steffen Nissen (Fast Artificial Neural Network Library)

Fast Artificial Neural Network Library

2-LWNN library by Peter van Rossum (Light Weight Neural Network Library)

SourceForge.net: Lightweight Neural Network

3-A good hardware acceleration for large computational application:

by using GPU multi processing threads (NVIDIA CUDA , AMD ATI...)

Nvidia is provided a good C library for using GPU power in your application

a comparision:

(Intel Core 2 Quad Extreme QX9770 : less than 100 GFLOPS (single 32bit float) or (double 64bit float) )

(NVIDIA GTX260 , GTX280: 933 GFLOPS(single 32bit float) and 100 GFLOPS (double 64bit float))

CUDA Zone -- The resource for CUDA developers

Is there anyone or company interested to build a CUDA enable Add-on for MetaTrader4 to speedup the calculation part of the platform such as optimization.

 
barnix:
I have started a C++ project: Better nn EA clone on google.

I want to make it free:

You can se the project at link:

Home (Better EA Clone)

You can find all C++ examples for designing a C++ command line tester.

For Better EA we must create a command line tester first in C++.

All project's it is in codeblocks.

On windows you must install Cygvin with boost libs.

hi barnix,very interesting with it,but seems there is something wrong with this project site,i can't visit it,could you check it?

 

The site is public.

Anyone in the world may view this site.

lvsefa:
hi barnix,very interesting with it,but seems there is something wrong with this project site,i can't visit it,could you check it?
 

Can u suggest the best EA, barnix

Barnix, just want to know and interest from Better EA Clone..

which want must to take... lets say the best...

please give idea Barnix..

 

Not hot on neural networks for this application.

Hi traders. I know I will not be particularly popular by saying this, but I don't think that developing trading methods with neural networks, at least at the level of knowledge I see them here, is likely to be long-term successful.

Before you hit "reply" with an angry message, please consider. You'll have to take this on faith, but I am a reasonable expert on machine learning & neural networks. I have a PhD in physics, and specialized in nonlinear data analysis. In my day job I create statistical models, including neural networks for a technology company which sells the software to banks--application is commercial banking not trading. For work I wrote a pretty state-of-the-art neural network trainer & analysis program in C++ over the last few years, which includes some features I haven't yet seen in the academic literature. I am not an expert in quant trading but I do have some mathematical and financial understanding.

Anyway, what I see now as the problem is the classic GIGO -- 'garbage in, garbage out'. Much more important than the statistical classifier/regression model (of which artificial neural networks are one popular form) is understanding of the fundamental data. What is most important is "what variables should I be using, are they reliably predictable, is my model stable, etc, etc."

My recommendation is simple: before going to a neural network model, you need to have a simple regression model working, i.e. capable of predicting the phenomenon in question with statistical confidence. I love complex nonlinear methods, but I'm recommending linear regression (logistic regression if you are predicting a categorical variable instead of a continuous variable).

Of course this ought to include significant variable selection, and very importantly cross-validation and regularization (i.e. "ridge regression"). You ought to think about the inputs and target variable, and try various nonlinear transformations of the input and output, i.e. manual ones.

Fitting a single regression model can be done in one deterministic step, unlike neural networks (problem is called "convex" in optimization-speak), but then you ought to be doing it over random resamples of the population and folds in cross-validation. At first, bypass the complexity and heuristics of fitting neural networks (yes, it really is a "black art" unfortunately) and work on more important data and statistical issues with a simple, linear model first.

A former colleague of mine did work in a quantitative hedge fund and he asserted that anything which worked had to work with simple regression. After that, then yes, fancier stuff might be able to extract a bit more performance, but basic research starts with simplicity.

I'd much rather trade a linear model with 4 inputs (possibly derived themselves with some fundamental insight) which was crafted from well bootstrapped and bagged models (look up "boosting" and "bagging") with enough out of sample analysis to convince me there really is some phenomenon there. Even in our products we can often get 80% of the performance of the neural network with regression models (there is lots of secret sauce in the input variables).

I really do like neural networks---in the right application they do kick ass.

 

Please Share Your Knowledge

mbkennel:
Hi traders. I know I will not be particularly popular by saying this, but I don't think that developing trading methods with neural networks, at least at the level of knowledge I see them here, is likely to be long-term successful.

Before you hit "reply" with an angry message, please consider. You'll have to take this on faith, but I am a reasonable expert on machine learning & neural networks. I have a PhD in physics, and specialized in nonlinear data analysis. In my day job I create statistical models, including neural networks for a technology company which sells the software to banks--application is commercial banking not trading. For work I wrote a pretty state-of-the-art neural network trainer & analysis program in C++ over the last few years, which includes some features I haven't yet seen in the academic literature. I am not an expert in quant trading but I do have some mathematical and financial understanding.

Anyway, what I see now as the problem is the classic GIGO -- 'garbage in, garbage out'. Much more important than the statistical classifier/regression model (of which artificial neural networks are one popular form) is understanding of the fundamental data. What is most important is "what variables should I be using, are they reliably predictable, is my model stable, etc, etc."

My recommendation is simple: before going to a neural network model, you need to have a simple regression model working, i.e. capable of predicting the phenomenon in question with statistical confidence. I love complex nonlinear methods, but I'm recommending linear regression (logistic regression if you are predicting a categorical variable instead of a continuous variable).

Of course this ought to include significant variable selection, and very importantly cross-validation and regularization (i.e. "ridge regression"). You ought to think about the inputs and target variable, and try various nonlinear transformations of the input and output, i.e. manual ones.

Fitting a single regression model can be done in one deterministic step, unlike neural networks (problem is called "convex" in optimization-speak), but then you ought to be doing it over random resamples of the population and folds in cross-validation. At first, bypass the complexity and heuristics of fitting neural networks (yes, it really is a "black art" unfortunately) and work on more important data and statistical issues with a simple, linear model first.

A former colleague of mine did work in a quantitative hedge fund and he asserted that anything which worked had to work with simple regression. After that, then yes, fancier stuff might be able to extract a bit more performance, but basic research starts with simplicity.

I'd much rather trade a linear model with 4 inputs (possibly derived themselves with some fundamental insight) which was crafted from well bootstrapped and bagged models (look up "boosting" and "bagging") with enough out of sample analysis to convince me there really is some phenomenon there. Even in our products we can often get 80% of the performance of the neural network with regression models (there is lots of secret sauce in the input variables).

I really do like neural networks---in the right application they do kick ass.

Hi mbkennel,

happy to see you here,

please share your knowledge on properly selection of Inputs and Outputs,

for preventing GIGO! ( a good description )

please provide some examples if possible.

regards.

OTR

 

Data mining tools

Hello,

has anybody tried to use free datamining tools like Knime or Rapidminer?

They have NNs, SVM and other learners/predictors built in ready to use.

Andy

 

Better EA results:

FOREX*PAMM

1...737475767778798081828384858687
New comment