I think many of them should exist in a Q&A place.
Thanks a lot for answering Popov
]]>I'll try to answer to your questions one by one
1. If "out of sample" parameter=In Sample, does "complete backrest criteria" ONLY matters for validation? (I think yes)
Yes, it is correct. In that case we don't have OOS zone and the complete backtest is "In Sample". It literally means that the backtester sees and uses the complete data set.
2. If "out of sample" parameter=10% OOS, does "In Sample" AND "Out of Sample criteria" ONLY matters for validation?
(I think yes)
when we use OOS, we have three zones: In Sample, Out of Sample and Complete Backtest. The program validates the strategy against all these three criteria sets.
3. If in Generator settings "The strategies fulfill the Acceptance Criteria" is not selected, WHY Generator strategies number is different than the passed Validation (inside Generator) Strategies Number; Is it correct? (it should be the same)
Even when the Acceptance Criteria are off the Generator has its own internal validation rules. The are simple:
- Strategy must be on profit (Profit > 0)
- Minimum 5 deals.
4. I have the same question for optimization Settings. (it should be the same number)
The optimizer accepts new strategy versions only if they are "better" than the original strategy. Here "better" means that the strategy has a higher value of the selected "Search best" param.
]]>1. If "out of sample" parameter=In Sample, does "complete backrest criteria" ONLY matters for validation? (I think yes)
2. If "out of sample" parameter=10% OOS, does "In Sample" AND "Out of Sample criteria" ONLY matters for validation?
(I think yes)
3. If in Generator settings "The strategies fulfill the Acceptance Criteria" is not selected, WHY Generator strategies number is different than the passed Validation (inside Generator) Strategies Number; Is it correct? (it should be the same)
4. I have the same question for optimization Settings. (it should be the same number)
GD
]]>