I know neural nets have been a great hope in the IT, but as far as I remember they were not able to be more than a kind of trainable logic-filter: A and B but not C and a bit of D...
Your EA generator is based a huge amount of patterns and indicators!
But isn't it trapped by over adapting just because of its big pattern-base?
Do you have and would you like to compare backtest results against forward tests?
You are right, to successfully results, neural nets need huge number of patterns, indicators and great quality, reliable algorithms, it's artificial intelligence - approach should be serious.
But, unfortunately, thousands of beginners not experienced programmers with the low quality algorithms forms an opinion about the ineffectiveness of neural networks.
Our company has several years of experience in the creation of reliable tools for fast and quality generation of neural network algorithms - indicators and EAs.
In order to make sure neural networks operate on the forward period, you can download our free indicators, their training was conducted at the beginning of the 2014 year.
To add comments, please log in or register