OptAtlas
Method

Lower Bounds

Bounding the optimum to prune exact search and certify solution quality.

Also called: Lower Bounds · lower bound · 하한값 · Martello-Toth bounds

The bounding technique that underpins both the efficiency and the quality guarantees of exact methods (classified as a method in v1). In 2D cutting & packing the bounds range from area bounds (total piece area ÷ bin area) to Martello–Toth-style discrete bounds; they let branch-and-bound prune unpromising branches and certify the optimality gap of a heuristic solution. Lower bounds for two-dimensional packing are surveyed by Lodi, Martello & Monaci (2002).

Claims & evidence

Every relationship is a claim with an evidence grade; applicable problem-to-problem relationships also carry an equivalence level. See the evidence policy.

RelationshipClaimEquiv.EvidenceSources
shares method withBranch and BoundLower-bound computation is the core component that drives pruning in branch-and-bound — the stronger the bound, the more of the search tree it cuts.E2A
  • ATwo-dimensional packing problems: A survey