Adleman's theorem

From Complexity
Revision as of 00:32, 13 August 2010 by Vipul (talk | contribs) (Created page with "{{complexity class containment| smaller = BPP| larger = P/poly}} ==Statement== ===In terms of complexity classes=== The complexity class BPP (bounded error probabilistic p...")
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

This article gives the statement, and proof, of one complexity class being contained in another one. The smaller complexity class is BPP and the larger complexity class is P/poly.

Statement

In terms of complexity classes

The complexity class BPP (bounded error probabilistic polynomial time -- polynomial time where the probability of error is bounded below ) is contained in the complexity class P/poly (languages where, for every length of input string, there is an advice string using which the algorithm runs in polynomial time).

Loose philosophical formulation

The statement is that:

Probabilistic algorithm that is independent of length Failed to parse (unknown function "\leadsto"): {\displaystyle \leadsto} Deterministic algorithm that depends on length

In other words, we can shed the probabilistic nature of the algorithm but must instead accept length-specific advice that shapes the algorithm.