Hi:

I ve worked with some softcomputing techniques for clustering, chaotic time series modelling. I know lot things, but there are extensive information, every day somebody create a new neural network architecture telling everybody that his neural network is the best.
Since i knew about forex i saw an interesting field for my knowledge and for some matlabs scripts made a long time, i tried predict price of EUR/USD using perceptrons but they suffer a lot of overfitting, memorize training data but predict like crap, althougth with some improvements with cross validation, local models make it better and are conceptually thousands times more simple than neural networks, but take a lot of computational time cause to create the model you need serach along all your data set nearest neighboors from your query point, in test sets it does make errors betwen 0.0018 and 0.0030 pips in data of 1h hour. (A local model is like a linear regression but you dont use all points, you only use the k-nearest to your point)
There's another class of neural networks created by russian investigators, they are polinomial neural networks, they have some few parameters, 'cause network it self is self organizative, algorithm it self decide number of nodes, polinomial function to use into the node and number of layer. An interesting work about polynomial nn in forex is here, is very good.

http://www.cs.uoi.gr/~foka/papers/foka_msc.pdf

I proved it with matlab but results are weird, i obtained a shifted version of prices like prediction, at first i thougth that i was wrong, but training an state of art neural network from neurosolutions i got the same. At last i found the very very, generalization of polinomial neural networks volterra kernels, this behaves like an infinite terms polynomial network (very complex concepts) is like a SVM(support vector machine), but i still follow getting the same shifting in predictions, this shifting make me errors in the same range of local models, if i would have to choose i would pass a local model to mq4 i still have the same results that with a very, very complex model like volterra kernel and better than perceptrons
I've read a lot about state of the art, there's two related concepts in neural networks that looks promising, but are still complex first are liquid state machines and others are echo
state networks, but i'm tired and searching more simple methods.

I thing that is better try to predict turning points using different indicator more than trying to predict price it self, trying price itself is very, very difficult (for not say nearly imposible still with state of art softcomputing techniques). Also, this strategy is very profitable like was demostrated in EA championships, my advice don't loose your time trying predict the price. I was coding differents methods in matlab in my free time during nearly three years(fortunatelly not all for forex), still the knowledge is very usefull and i've learnt sooooooooooooo much, but i insist don't loose your time trying predict the price.


See ya guys

I wonder what kind of software is used to do this prediction.Can anyone help me? http://neuralforex.wordpress.com

I am starting to make my experiments here. Unlikely I am from a different background. My degree is Artificial Intelligence so more technical towards AI techniques such as NN and more newbie in forex unlike the rest of the guru's in this forum...

using the historical price to Predict the future price it self is kinda impossible due to many factors..

as a sum up from the previous experiences.. whats are the good inputs and there expected outputs for a NN ..

anyways,, whats your opinion about whats the good values/indicators to be used as inputs in Neural Network and what would be there expected output(indicators, curve movements..etc..)..

I will test it on a series and different types of neural networks with back error probagation. also found interesting publication about genetic mutation for the results. I just have some problem now on where to find reliable data to feed the NN... where do you use your historical data .. any sources?


Regards,
Hidden Ox.

prediciting price is surely impossible..
but what if we implement fuzzy logic instead of precision prediction
i.e just predict either it going up or down..
is it applicable?
cos even when using indicator it still not exactly right , right?
it's game of probabilities..

yupss fuzzy logic can be used to determine what's price current condition , it's use statistical strategy but more better use with NN to recognize the pattern

===================
Forex Indicators Collection

Dear,

I can build a C++ DLL that you can use as your link between your neurosolution and your mql4 script. messege me if you still interested.

I think everyone would be interested in that, I've tried to get a dll to work with Matlab's neural network toolbox and it's not easy to do. The toolbox trains the net with a levenberg-marquardt algorithm, and I would like to see some out-of-sample performance statistics...
Neurosolutions would be an equally performant tool to do this I think.

__________________
tensigma.blogspot.com

Hello Coders,

i'm working on an multi-agent-system. it evolves with the help of thousend agents and some genetic programming trading strategies. but in mql4 it is very lame. are you interested in? i need some expirience coders, which like to code a dll as interface between mt4 and the MAS.

maybe it is a little bit off-topic but it is AI stuff.

i made a prototype ea which initialise the genomes. my problem is the logic with tree-based structures in mql4. because it is type depentend. sorry for my bad english. but it is a heavy topic to explain right. maybe XML would help, but do not know to implement it.

Thanks in advance

PriNova

This is a GREAT thread, forex tsd rocks!!!

I am a final year computer science student,

As my end of year project, I am creating a web-based charting software (very similar to the Yahoo Finance Charts with all the usual indicators i.e. Stochastics, Moving averages, RSI, etc...)
but with a major addition - the neural network based indicator.


As freely available neural network indicators seem rare, I would like to know:

- if anybody on these forums would consider using such a system? (my charting solution)
- do you have any knowledge about neural networks? (none required)
- any suggestions? or features you would like to see?


Note:
- The aNN (artificial neural network) indicator will be displayed using default setting values (number of hidden neurons, etc...) so that even users with NO knowledge of Neural Networks can use the indicator.
- The aNN indicator will also allow a user to change these default values (for advanced/experimenting users).
- Stock market feeds will be used at first. But FOREX currency pairs data can easily be added.

All comments welcome