> I often use brute force after having used D&C to 'hone in' on a 'juicy' > part of the parameter space, but one that is not necessarily the best part > of the whole space. You can sometimes find better parameters by being more > thorough than D&C is in the initial passes. > >
I use a similar approach. My first pass is D&C to find "interesting" regions, and then BF for a more thorough search in these regions. I also often use D&C with its most dense coverage setting, for the overnight optimization runs. With the high coverage density, the D&C accuracy approaches that of BF, and the advantage is that I don't need multiple BF runs with the ever smaller step sizes. The D&C has improved a lot over the last few releases, and I am quite happy with it. I am yet to see an instance where it missed the best parameter set, even with the relatively sparse coverage setting. --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "JBookTrader" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [email protected] For more options, visit this group at http://groups.google.com/group/jbooktrader?hl=en -~----------~----~----~----~------~----~------~--~---
