On-Demand Hard Instances for Algorithm Evaluation
In brief
- Automated method to generate hard instances by optimizing solver runtime.
- Tested on four NP-hard problems: SAT, TSP, Hamiltonian cycle, and flow-shop scheduling.
- Generated instances outperform random and classic benchmarks in difficulty.
A method optimizes instance generation to stress-test NP-hard algorithms. It produces cases far more challenging than random draws or curated benchmarks by analyzing solver runtime. Validated on four problems (SAT, TSP, Hamiltonian cycle, flow-shop scheduling), it uncovers known structures and outperforms existing references.
This summary gives you the essentials. The full article, its figures and quotes are on cs.AI updates on arXiv.org.
Read the original articleAI-generated summary of the source, checked by rules. Report an error