Approximation-preserving reduction

Approximation-preserving reduction (APR) is a concept in computational complexity theory and optimization that relates to how problems can be transformed into one another while preserving the quality of approximate solutions. It is particularly useful in the study of NP-hard problems and their approximability.