Best Neural Network Software to Begin ?

[Gillou 10]

all is in the title.

 

free or not

 

working with MC

 

thanks for yourreply

[metamp 10]

FANN is the only open source NN I know the FANN DLL library can be used in advanced trading pratforms like Neo Ticker or Ninja

 

Fast Artificial Neural Network Library | Download Fast Artificial Neural Network Library software for free at SourceForge.net

 

Just to let you know Neural network doesn't guarantee or being stright doean't solve the problem of finding a robust approach

one thing it does very well is finds the local minima just like GA optimizers do, and trust me every data set no matter how large has it's properties and your NN will find them

by implementing an NN to your system you are using the most complex solution of the problem

out of all possible solutions of a "PROBLEM" or challenge if you like the best is the most simple one

this has been proven many times

 

many newcomers think that if they get an advanced software or use an NN they will succeed, the reality is that there is no way to predict the future and the history doesn't repeat itself, there is no such a thing as a robust system

 

hope this helps

[Gillou 10]

Many thanks Metamp

[ramora 10]

You might also want to look at snns. I used it a lot several years ago and found it very complete and well documented. The TDNN net was very powerful. I had purchased several commercial packages and ended up using SNNS instead.

 

And it is free.

 

Good luck.

 

ramora

 

Stuttgart Neural Network Simulator

 

SNNS (Stuttgart Neural Network Simulator) is a software simulator for neural networks on Unix workstations developed at the Institute for Parallel and Distributed High Performance Systems (IPVR) at the University of Stuttgart. The goal of the SNNS project is to create an efficient and flexible simulation environment for research on and application of neural nets.

 

The SNNS simulator consists of two main components:

 

1) simultor kernel written in C

2) graphical user interface under X11R4 or X11R5

 

The simulator kernel operates on the internal network data structures of the neural nets and performs all operations of learning and recall. It can also be used without the other parts as a C program embedded in custom applications. It supports arbitrary network topologies and, like RCS, supports the concept of sites. SNNS can be extended by the user with user defined activation functions, output functions, site functions and learning procedures, which are written as simple C programs and linked to the simulator kernel.

 

Currently the following network architectures and learning procedures are included:

 

* Backpropagation (BP) for feedforward networks

o vanilla (online) BP

o BP with momentum term and flat spot elimination

o batch BP

* Counterpropagation

* Quickprop

* Backpercolation 1

* RProp

* Generalized radial basis functions (RBF)

* ART1

* ART2

* ARTMAP

* Cascade Correlation

* Recurrent Cascade Correlation

* Dynamic LVQ

* Backpropagation through time (for recurrent networks)

* Quickprop through time (for recurrent networks)

* Self-organizing maps (Kohonen maps)

* TDNN (time-delay networks) with Backpropagation

* Jordan networks

* Elman networks and extended hierarchical Elman networks

* Associative Memory

[UrmaBlume 10]

For us NNs are usually not enough to do the whole job.

 

The most effective models we have built come from a cascade of technologies. An example of such a cascade would be a model developed by first optimizing input preprocessing and input selection with genetic algorithms.

 

Then input data is used to develop a model using technologies that develop their models through what is called feed back mathematics, machine learning. Neural Networks and certain regression packages are used for this phase. Genetic algorithms can be used in this process to select and optimize such parameters and the number of nodes in middle layers of NNs and the transfer fuctions that weigh and pass the data between them. Our most used inter-nodal transfer function is the long-tailed sigmoid.

 

After the model is developed its output can be further optimized using decision trees or rules generators.

 

This site is a good place to find articles on and vedors of such tools. Every year that have a competition that ranks tools according to class.

 

For genetic optimizations we have written our own.

 

For the modeling process we like MARS® (Multivariate Adaptive Regression Splines) from Salford Systems in San Diego. We have built an application that converts the final MARS function to Trade Station's Easy Language so that we can deploy these models so that they can make their predictions in real time with just a cut and paste.

 

For rules generation we use CART, again from Salford Systems described on their website as "a robust, easy-to-use decision tree that automatically sifts large, complex databases, searching for and isolating significant patterns and relationships." We also use WizWhy from WizSoft. WizWhy cost around $4k and has served us very well.

 

Our process is described here which is an update and rewrite of this thread here on TL.

 

Good Luck

 

UB

[bomberone1 10]

Folks I am seraching a neural networks coded in easylanugage or that is it possible to link to multicarts by a mouse click, are there?

[Cortex7 10]

With regard to the original post, I am currently in the process of developing a charting/trading application (Cortex7) that uses Artificial Neural Networks (ANNs) to generate buy/sell signals.

 

The charting application will be freeware, if you want ANNs enabled then a license key must be entered.

 

At this stage the networks are trained to predict if tomorrows close will be higher than today's. Testing shows that the ANNs perform well. Testing involves training the network on all available data excluding X number of recent data points. Then a trade test is performed on the latest X data points.

 

Interestingly I have found that different ticker symbols have very different ratios of order/disorder, some symbols can be predicted far better than others. And different symbols need different ANN topologies for optimal prediction.

[Mr_You 10]

I haven't used it, but I just came across "Encog" which has instructions for integration with NinjaTrader. I'm guessing integrating with other platforms would be similar.

 

Development appears to be active, v3 is in beta testing.

[bomberone1 10]

is it possible to do integration into multicharts?

[Cortex7 10]

I forgot to post a link to cortex7:

Cortex7: Artificial Neural Network Stock Charting Software