Reply

Is Random Generation better to avoid Curve Fitting

2 replies

rafaeldelrey

Customer, bbp_participant, community, 18 replies.

Visit profile

7 years ago #115577

Hi,

 

I have been using SQ for a while, and trying lots of combinations of workflows.  Considering the Genetic Generation uses a evolution (which is a kind of optimization), would you say the Random Generation is less prone to create Curve Fitted strategies?

 

 

0

Edinho

Customer, bbp_participant, community, 36 replies.

Visit profile

7 years ago #139481

From my experience is 50/50 I have strategies generated both ways and they are making some profits. You need to realize no strategy will be profitable all the time.You need to diversify strategies across as many markets as possible.From my experience, good markets are GOLD , EURUSD , AUDJPY and DAX .Also if you are investor with larger investment I would definitely recommend trading the options as well.

0

Karish

Customer, bbp_participant, community, sq-ultimate, 443 replies.

Visit profile

7 years ago #139484

From my experience is 50/50 I have strategies generated both ways and they are making some profits. You need to realize no strategy will be profitable all the time.You need to diversify strategies across as many markets as possible.From my experience, good markets are GOLD , EURUSD , AUDJPY and DAX .Also if you are investor with larger investment I would definitely recommend trading the options as well.

Great answer, i will add that you can also diversify on only 1 pair if you want to,

as for a newbie go with EURUSD pair this is the most usable pair, again, for diversification you do not have to use many markets..,

 

as for the main question about Genetic / Random generations..,
Genetic could lead to more curve-fitted strategies then Random Generation,

 

but that depends on the user.., always use simple rules..,

i would go with Random Generation always because that way the system will generate Random rules and not Genetically modify the already exciting, but again.. depends on the user, some prefer Genetic more.

0

Viewing 2 replies - 1 through 2 (of 2 total)