READ-ONLY

The forum is now a read-only archive.

For bug reports & platform questions → [email protected]

Our community lives on Discord and YouTube — come join us!

Questions about optimization

6 replies

Fluke

Subscriber, bbp_participant, community, 41 replies.

Visit profile

11 years ago #113921

Hi, I’m almost at the end of the trial period, I would understand, as I will optimize strategies? I understand that I walk forward matrix suggests the optimal combination of runs oos%.
Where should I use these runs oos% combinations?
I have to use on the walk forward optimization?
If so, how do I get the source code with the parameters optimized?

0

Fluke

Subscriber, bbp_participant, community, 41 replies.

Visit profile

11 years ago #131208

in other words, when I have to re-optimize, I can only use simple optimization, (which generates new strategies visible tab optimization result), or am I that I can not make out new setup using results walk-forward ?

The source code of Walk forward optimization strategy, is exactly the original strategies, and so my question is, why strategyquant not applicate the optimization parameters? it’s my error? where i’m wrong?

0

Fluke

Subscriber, bbp_participant, community, 41 replies.

Visit profile

11 years ago #131228

Ok, with help of old pdf manual of StrategyQuant i have understand the think behind, walk-forward optimization and walk-formward matrix.

Walk-forward matrix: i use this to find the best combination of runs and oos% for my reoptimize.
Walk-Forward Optimization: Use this to valutate how the system involves in the real market when we reoptimize periodically the strategy.

Simple optimization: i use it to reoptimize periodically the strategy.

Now i have another question, why the software allow reoptimize in simple mode (for all the data range selected), and not allow to create a strategy reoptimized with walk forward mode? In other words, why not allow to generate strategy with parameters founded whit walk-forward test? Thank to Mark for the answer, meanwhile that look the answer i wish you good job  😀

0

Heilpraktiker

Customer, bbp_participant, community, 46 replies.

Visit profile

10 years ago #136612

Ok, with help of old pdf manual of StrategyQuant i have understand the think behind, walk-forward optimization and walk-formward matrix.

Walk-forward matrix: i use this to find the best combination of runs and oos% for my reoptimize.
Walk-Forward Optimization: Use this to valutate how the system involves in the real market when we reoptimize periodically the strategy.

Simple optimization: i use it to reoptimize periodically the strategy.

Now i have another question, why the software allow reoptimize in simple mode (for all the data range selected), and not allow to create a strategy reoptimized with walk forward mode? In other words, why not allow to generate strategy with parameters founded whit walk-forward test? Thank to Mark for the answer, meanwhile that look the answer i wish you good job  😀

This i want to know, too!

0

Tomas Matejka

Administrator, sq-ultimate, 2 replies.

Visit profile

10 years ago #136631

Analysis of walk-forward results can tell you if the strategy is a good candidate for periodic reoptimization. WFO tells you whether the strategy is robust enough and if its performance can be improved by re-optimization. It does not tell which is the best parameter setting. This is the job for the Simple optimization process. That’s why you cannot save WFO results as a strategy file.

0

mabi

Customer, bbp_participant, community, sq-ultimate, 252 replies.

Visit profile

10 years ago #136633

Analysis of walk-forward results can tell you if the strategy is a good candidate for periodic reoptimization. WFO tells you whether the strategy is robust enough and if its performance can be improved by re-optimization. It does not tell which is the best parameter setting. This is the job for the Simple optimization process. That’s why you cannot save WFO results as a strategy file.

 

SQ team should spend some time making explainatory short video series of the functionality of SQ. Maybe some webinars even.  

0

Heilpraktiker

Customer, bbp_participant, community, 46 replies.

Visit profile

10 years ago #136670

Analysis of walk-forward results can tell you if the strategy is a good candidate for periodic reoptimization. WFO tells you whether the strategy is robust enough and if its performance can be improved by re-optimization. It does not tell which is the best parameter setting. This is the job for the Simple optimization process. That’s why you cannot save WFO results as a strategy file.

Thank you, Thomas 🙂

0

Viewing 6 replies - 1 through 6 (of 6 total)