28. Additional Applications in Other Settings
David Peleg · Society for Industrial and Applied Mathematics eBooks · 2000
The basic premise behind employing LP-representations is that using them (as a substitute for using the entire network) allows us to maintain smaller amounts of information (namely, remember fewer edges, know less about the structure of the graph and so on), thus reducing the computational and storage costs involved. The price to be paid is that using sparse, approximate representations incurs a certain loss of accuracy. Hence while our focus in this book was on applications in the area of distributed network algorithms, it is evident that these representations should in principle be useful in any situation where exact solutions are not essential or are too expensive to obtain.