Optimise traders with coevolution hill climbing gradient with random restarts

Taketrung / betting-ai

Automatically exported from code.google.com/p/betting-ai

1 stars 1 forks source link

Optimise traders with coevolution hill climbing gradient with random restarts #99

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago

1) Look around randomly for trader with positive expected profit, 
2) If one is found then optimise it with co-evolution hill climbing gradient 
algorithm. 
3) When local maximum is reached then go back to point 1 until total number of 
iterations is reached.

Original issue reported on code.google.com by daniel.k...@gmail.com on 4 Mar 2011 at 6:05

GoogleCodeExporter commented 9 years ago

Original comment by daniel.k...@gmail.com on 4 Mar 2011 at 6:06

Added labels: Priority-Critical
Removed labels: Priority-Medium

GoogleCodeExporter commented 9 years ago

Original comment by daniel.k...@gmail.com on 4 Mar 2011 at 6:09

Changed state: Started

GoogleCodeExporter commented 9 years ago

TODO:
Pass Solution[T <: ITrader] to HillClimbing instead of just Itrader, to not 
lose possibly the best initial solution.

LocalOptimisation currently last for n number of iterations. Develop a bit more 
clever stop criteria, e.g. terminate optimisation if solution is not improving 
for a given number of iteration.

Original comment by daniel.k...@gmail.com on 23 Mar 2011 at 7:25

GoogleCodeExporter commented 9 years ago

Original comment by daniel.k...@gmail.com on 14 Apr 2011 at 11:18

Changed state: Verified