Sublinear-time reductions for big data computing

The inefficiency of polynomial-time ( PTIME ) algorithms in the context big data indicates a departure from traditional view on tractability. In recent years, sublinear-time have been regarded as new standard for tractability computing problems. Based prior work complexity classes [1] , this paper focuses designing appropriate reductions specialized particular, series pseudo-sublinear-time are proposed, and their properties systematically investigated. It is proved that PsT properly contained P any -complete problem PsPL i not must be witness ∖ NC . Several examples also given to illustrate usefulness proposed reductions. Then, cope with problems intractable sublinear time even after preprocessing, L -reduction extended pseudo-sublinear time. Finally, it there no PPL under DLOGTIME reduction. • Four kinds studied. all can solved preprocessing. if PsT-complete or PsPLi-complete P- complete, then ≠NC. PPL-complete