This gives a forex trader mega strategy which works by backtesting thousands of standard combination techniques from a library of strategies / indicators.
It then chooses to open margin trades in real time based on indications from combinations with a sufficient ratio of winning test trade confirmations.
It uses a growing list of methods to compute suitable position "stop loss" / "take profit" bounds based on percentages and computed pivot points (like demark, fibonacci retracement, etc).
To build the mega strategy, it employs tons of indicators from joeldg's repo.. I added a number of tools for backtesting and evaluating across a range indicators as well as a tools for evaluating the success of the "Terry Tibbs mega strategy".
I'm currently working on training some tensorflow models to utilise this vast set of indicator data + some moving averages, to predict the likely success over a range of simple margin position builder techniques.
Much love!