Narrow search space
Optimize one hypothesis layer at a time.
The optimizer explores parameter surfaces, compares multiple objectives and preserves the selected candidate's neighbourhood beside the cost of search.
Search, rank and freeze a candidate
Optimize one hypothesis layer at a time.
Prefer stable parameter regions over an isolated peak.
Keep final evaluation separate from selection.
Record every attempt and reflect it through DSR and PBO.
Plateau analysis checks neighbouring values. WFO tests repeated optimisation across time. DSR adjusts Sharpe for all trials. These tests complement rather than replace one another.
Inspect the parameter neighbourhood
Run WFO across periods
Apply DSR/PBO to every trial
Selected parameter region
Historical, statistical and model results do not promise future returns. kquant provides computational tools and technical execution, but does not assess whether a specific trade is suitable for a user.