The Spinal Implant EA is an unsupervised Neural Net Scalper with Fractal Geometry filtering. This is an implementation of unsupervised neural net with Genetic algorithm optimization by MT4. It has the following characteristics.
The file and the discussion are available only for registered users. Register for Free.
1. Signals. The neural net provide output signals. The general interpretation of these signals will be some sort of "binary" classification, like "buy" (>0) or "sell" (< 0). I simply insert them in the Fractal scalper EA trading strategy as a rule (condition) to buy and sell.
In the fractal scalper there is another condition that is that the estimation of Hurst exponent (IVAR,FGDI) is large enough. Those parameters can be also optmized.
2. Unsupervised. This neural net is "unsupervised" neural network, and that is meaning that it does not need to be trained by showing the correct answers, like most neural networks (i.e., "supervised" neural networks). You do not teach them by providing any kind of actual output which they learn to reproduce, as with most neural nets.
3. Evolutionary. The Neural net "learns" how to give its signals based upon evolutionary pressure provided in the native MT4 algorythm. The genetic algorithm (GA) in the MT4 "evolves" the neural net that give better and better signals. Survival of the fittest controls the evolutionary process as usual, where fitness is determined by how much money the NN makes (you can use what to optimize in the MT4, e.g. Balance, Profit Factor etc.).
There are two other mods of the Spinal implant EA.
ОтговорИзтриванеSpinal Implant Sigm EA with Sigmoidal transfer function
Spinal Implant Radial basis NN EA with Gaussian transfer function
It is a matter of choice and experimentation. The basic idea is that you can implement the available perceptron functions in other EAs.
ОтговорИзтриванеYou need to consider the Elliotware approach and the market state analysis in the decision of the training range.
ОтговорИзтриванеhttp://beathespread.com/groups/profile/77/elliottware-machine-learning-extension-of-elliott-wave-principle
http://beathespread.com/groups/profile/436/trend-or-range-and-beyond-it-market-state-analysis
I want to mention that this basic neural net implementation is in active development, so multiple new versions have been implemented. Every version has its own characteristics. The best implementations were those profiting with the complex representation see this article.
ОтговорИзтриванеThis new design completely outperformed out of sample the normal neural net design.
http://www.fxhackers.com/2013/01/chaos-and-complex-representations-for.html