Self optimising robot
Created at 26 Nov 2018, 00:32
Self optimising robot
26 Nov 2018, 00:32
Short question...Is it possible to code a robot that you can code to be able to optimise its self. If so how would you go about doinng this?
afhacker
26 Nov 2018, 05:55
Yes its possible and there are different ways to do it, one simple approach will be looking back to history, suppose your bot made 10 trades and after closing the 10th trade your bot will analyze the trades result and it will find out the profitable trades characteristics like in what time of day your bot performed will, in what kind of market (trending or consolidation) it did well,... etc
After that you will use those extra filters on 10 future trades, then you analyze the result of applying those new filters in your 10 future trades if those filters helped you keep applying those filters if didn't you get rid of those filters.
And you keep repeating the above step after every 10 trades, for analyzing you can use the past x trades or all of the past trades.
That was a simple way to do optimization inside your bot but there are other ways like using ML regression models and keep refitting the model after x number of days or weeks.
@afhacker