1996 Reports

Computational complexity of continuous problems

Wozniakowski, Henryk

Computational complexity studies the intrinsic difficulty of solving mathematically posed problems. Information based complexity is a branch of computational complexity that deals with continuous problems defined on spaces of multivariate functions. For such problems only approximate solutions are possible to compute. The complexity is defined as the minimal cost needed to compute an approximation with error at most. Error and cost can be defined in different settings such as the worst case average case probabilistic or randomized settings.