Hello, 

   I started some experiments with Fuzzy Logic in QuantConnect / Lean.

   Fuzzy Logic makes sense in technical trading, as the variables are considered fuzzy (many-valued logic) with "truth values" between 0 and 1. This is much more adequate for leading with subjective decisions and uncertainity, rather than with crisp values wich are "idealistic". Example, is RSI > 80 High ? What about 89 ? Not high at all ? :)
   Please, find an intro to fuzzy logic in the context of automated trading here

   In this example i used two classic indicators - RSI and Momentum - just for illustration purpose. 
      Many rules can run at the same time, and Linguistic Variables can have several memberships. 
    I am using the Accord.NET frammework Fuzzy Library (installed on LEAN since Machine Learning also from accord was installed):

http://www.aforgenet.com/articles/fuzzy_computing_basics/

Documentation:

http://accord-framework.net/docs/html/N_AForge_Fuzzy.htm

Feel free to contact me, for questions, or ideation sharing.
Have fun

 

Author