Flag to determine if the best solutions should be updated regardless of their quality
|Reported by:||mkommend||Owned by:||gkronber|
Currently the symbolic data analysis analyzers that track the best training or validation solution only update these, if the previous best quality is further improved. However, if a stochastic fitness function (e.g., sampling, sliding window, ...) is used, this approach is not feasible as the quality might change from evaluation to evaluation and the best solution should be updated regardless of previously found solutions.
Change History (5)
comment:3 Changed 4 years ago by mkommend
- Owner changed from mkommend to gkronber
- Status changed from accepted to reviewing