How does Random Generation work exactly?

Hi,

I’m curious. How does Random Generation work exactly? Is it completely random or does it eliminate the combinations that don’t work?

Say, for example, if it comes up with an ABC strategy and the strategy doesn’t pass the filters, will this exact same strategy (same indicators, rules, and parameters and everything) have a chance to appear again later within the same building session (as long as I don’t press the Stop button)?

in this mode StrategyQuant continually generates and tests new random strategies, one after another, until it is stopped. The top candidates (based on predefined criteria) are stored into Databank so you can review them later.

Pros:

• faster and simpler than genetic evolution
• it can run until it is stopped, so if you let it run for a few days it can generate and evaluate millions of strategies
• less prone to over-fitting, strategies is not further optimized or improved

Cons:

• once the strategy is generated it is not further evolved or optimized – but you can always use it in an initial population for the next build based on genetic evolution

Yes, I understand that. I’m just wondering whether the exact same strategy (same building blocks and same parameters) will have a chance of being generated again within the same building session.

Say, I run build strategies in the random generation mode for 6 hours and it generates Strategy ABC, will the Strategy ABC have a chance of being generated again within that 6 hours? Or will SQX go “oh, I’ve already generated this strategy, let’s remember that and don’t generate using these parameters again”?

I’m thinking about whether I should quickly generate a few hundred strategies and then run them through robustness tests or wait longer until I have around 5000 strategies before starting robustness tests.

Because, if the Random Generation mode keeps in memory and doesn’t generate duplicates, then it might be better to just wait until I have a bunch of strategies generated before running them through robustness tests.