You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
When comparing algorithms on "Instances Closed" plots, it is possible that some closed instances will not be attributed to any algorithm.
For example: an optimal method claims a best known bound equal to C, but it does not claim any plan. Another solver claims a best known solution with cost equal to C, but it does not claim any lower-bound. The instance is closed, yet according to the current filtering method, it does not appear in this plot. The problem arises because we only consider an instance closed if a submission reports a solution cost and a matching lower bound value.
How should we handle/attribute instances closed by a pair of algorithms working together? Suggest that we need an "Other" category in the corresponding plots.
The text was updated successfully, but these errors were encountered:
Arising from Issue #13.
When comparing algorithms on "Instances Closed" plots, it is possible that some closed instances will not be attributed to any algorithm.
For example: an optimal method claims a best known bound equal to C, but it does not claim any plan. Another solver claims a best known solution with cost equal to C, but it does not claim any lower-bound. The instance is closed, yet according to the current filtering method, it does not appear in this plot. The problem arises because we only consider an instance closed if a submission reports a solution cost and a matching lower bound value.
How should we handle/attribute instances closed by a pair of algorithms working together? Suggest that we need an "Other" category in the corresponding plots.
The text was updated successfully, but these errors were encountered: