Refine
Year of publication
Document Type
Keywords
- 3-Satisfiability (1)
- Algebraic models of computation (1)
- Algebraic proofs (1)
- BSS machines over C (1)
- Complexity theory Saturation (1)
- Decision problems for automata over uncountable structures (1)
- Exponential-time computations (1)
- Generalized finite automata (1)
- Hilbert Nullstellensatz (1)
- Linear and quadratic programming (1)
- NP-completeness (1)
- PCP theorem for NP over the reals (1)
- Probabilistically checkable proofs (1)
- Probabilistically checkable proofs Real and complex number computations PCP theorem Hilbert Nullstellensatz decision problem (1)
- Real and complex number computations (1)
- Real model of computation (1)
- Real number model of computation (1)
- Sparseness (1)
- Structure of complexity classes (1)
- Testing trigonometric polynomials (1)
- complexity (1)
Institute
We propose the study of query languages for databases involving real numbers as data (called real number databases in the sequel). As main new aspect our approach is based on real number complexity theory as introduced in [8] and descriptive complexity for the latter developed in [17]. Using this formal framework a uniform treatment of query languages for such databases is obtained. Precise results about both the data- and the expression-complexity of several such query languages are proved. More explicitly, relying on descriptive complexity theory over ℝ gives the possibility to derive a hierarchy of complete languages for most of the important real number complexity classes. A clear correspondence between different logics and such complexity classes is established. In particular, it is possible to formalize queries involving in a uniform manner real spaces of different dimensions. This can be done in such a way that the logical description exactly reflects the computational complexity of a query. The latter might circumvent a problem appearing in some of the former approaches dealing with semi-algebraic databases (see [20] , [18]), where the use of first-order logic over real-closed fields can imply inefficiency as soon as the dimension of the underlying real space is not fixed - no matter whether the query under consideration is easy to compute or not.
Regarding real number models of computation may be a helpful way to get deeper insight into the classical theory over Z. Therefore it seems useful to study the complexity of classical problems in the real model. In this connection, the problem of deciding the existence of a nonnegative zero for certain polynomials plays an important part because a lot of NP-problems over Z can be polynomial reduced to it.
The computational model of Blum, Shub, and Smale (1989, Bull. Amer. Math. Soc.21, 1-46) yields a numerous number of different algorithm-classes if the allowed set of instructions is varied. For two such classes we consider the "P = NP?" problem as well as the question of whether the NP- problems are decidable.
When comparing complexity theory over the ring Z (Turing machine) on one hand and over the ring R (Blum-Shub-Smale machine) on the other hand, it will be important to study how methods and ideas of the first can be transformed to the second one. In this sense the present paper is concerned with the relation between a characterization of the P = NP?-question for the Z-case (given by Krentel) and a special class of quadratic-programming problems which are important in the real model.
The complexity of linearly constrained (nonconvex) quadratic programming is analyzed within the framework of real number models, namely the one of Blum, Shub, and Smale and its modification recently introduced by Koiran (“weak BSS-model”). In particular we show that this problem is not NP-complete in the Koiran setting. Applications to the (full) BSS-model are discussed.
Relations between discrete and continuous complexity models are considered. The present paper is devoted to combine both models. In particular we analyze the 3-Satisfiability problem. The existence of fast decision procedures for this problem over the reals is examined based on certain conditions on the discrete setting. Moreover we study the behaviour of exponential time computations over the reals depending on the real complexity of 3-Satisfiability. This will be done using tools from complexity theory over the integers.
Shub (in "From Topology to Computation: Proceedings of the Smalefest." (M. Hirsch, J. Marsden, and M. Shub, Eds.), pp. 281-301, Springer-Verlag, New York/Berlin, 1993) proposes to attack the complex PFull-size image (<1 K) versus the NPFull-size image (<1 K) problem by focussing on lower bounds on testing the resultant of quadratic forms for zero. Taking up this question we show in the present paper a lower bound of order n3 for the resultant.
The PCP theorem has recently been shown to hold as well in the real number model of Blum, Shub, and Smale (Baartse and Meer, 2015). The proof given there structurally closely follows the proof of the original PCP theorem by Dinur (2007). In this paper we show that the theorem also can be derived using algebraic techniques similar to those employed by Arora et al. (Arora et al., 1998; Arora and Safra, 1998) in the first proof of the PCP theorem. This needs considerable additional efforts. Due to severe problems when using low degree algebraic polynomials over the reals as codewords for one of the verifiers to be constructed, we work with certain trigonometric polynomials. This entails the necessity to design new segmentation procedures in order to obtain well structured real verifiers appropriate for applying the classical technique of verifier composition.
We believe that designing as well an algebraic proof for the real PCP theorem on one side leads to interesting questions in real number complexity theory and on the other sheds light on which ingredients are necessary in order to prove an important result as the PCP theorem in different computational models.