Hi Juan,

I try to understand the concepts of what you are trying to obtain with this machine learning, could you explain this ?

Normally you develop a strategy with fixed settings (eg like a Macd crossing strategy) based upon past data, and assuming the future movements of price are relatively similar, and you let run your trading-algo.

With this machine learning concept, are you trying to further optimize the following of past changes in price movements, then with those optimized findings approaching the future price movements with the new strategy ?

Or when running a algo-strategy, with this machine learning concept, it adjust the setting of the variables automatically given the structure of price movements ?

(I hope my question is clear,  in this case it would be great to discuss this kind of topics face to face in a meeting with others, to avoid misunderstandings)