Graduate Texts in Mathematics: Difference between revisions
No edit summary |
|||
Line 1: | Line 1: | ||
== | {{Expert-subject|Computing|talk=Copy edit not needed|date=January 2012}} | ||
In [[descriptive complexity]], a branch of [[Computational complexity]] ''FO'' is a [[complexity class]] of structures which can be recognized by formulas of [[first-order logic]]. It is the foundation of the field of [[descriptive complexity]] and is equal to the complexity class [[AC0|AC<sup>0</sup>]]. Descriptive complexity uses [[First-order logic]] formalism, but does not use most of the usual notion associated to logic as proof theory or axiomatization. | |||
When the predicates are restricted to a set ''X'', ''FO[X]'' correspond to some other class, and there exists specific sets ''X'' such ''FO[X]'' has some interesting properties. In particular, ''FO[<]'' is equal to the set of [[Star-free language]]s. Having two different definition, in term of logic and in term of regular expression, is a hint showing that this class may have some mathematical interest, and that we may use together methods for both domain to do proofs. | |||
Similarly, various extensions of FO, formed by the addition of certain operators, give rise to other well-known complexity classes,<ref>N. Immerman ''Descriptive complexity'' (1999 Springer)</ref> allowing the complexity of some problems to be proven without having to look at them as [[algorithm]] problem. | |||
== Definition and examples == | |||
=== The idea === | |||
When we use the logic formalism to describe a computational problem, the input is a finite structure, and the elements of that structure are the [[domain of discourse]]. Usually the input is either a string (of bits or over an alphabet) the elements of which are positions of the string, or a graph of which the elements are vertices. The length of the input will be measured by the size of the respective structure. | |||
Whatever the structure is, we can assume that there are relations that can be tested, for example "<math>E(x,y)</math> is true [[iff]] there is an edge from <math>x</math> to <math>y</math>" (in case of the structure being a graph), or "<math>P(n)</math> is true [[iff]] the <math>n</math>th letter of the string is 1." These relations are the predicates for the first-order logic system. We also have constants, which are special elements of the respective structure, for example if we want to check reachability in a graph, we will have to choose two constants s (start) and t (terminal). | |||
In descriptive complexity theory we almost always suppose that there is a total order over the elements and that we can check equality between elements. This lets us consider elements as numbers: the element <math>x</math> represents the number <math>n</math> iff there are <math>(n-1)</math> elements <math>y</math> with <math>y<x</math>. Thanks to this we also may have the primitive predicate "bit", where <math>bit(x,k)</math> is true if only the <math>k</math>th bit of <math>x</math> is 1. (We can replace addition and multiplication by ternary relations such that <math>plus(x,y,z)</math> is true iff <math>x+y=z</math> and <math>times(x,y,z)</math> is true iff <math>x*y=z</math>). | |||
=== Formally === | |||
Let ''X'' be a set of predicate. The language ''FO[X]'' is defined as the closure by conjunction ( <math>\wedge</math>), negation (<math>\neg</math>) and universal quantification (<math>\forall</math>) over elements of the structures. We also often use existential quantification (<math>\exists</math>) and disjunction (<math>\vee</math>) but those can be defined by means of the first 3 symbols. The base case being the predicates of ''X'' applied to some variables. We always implicitely have a predicate <math>P_a(x)</math> for each letter <math>a</math> of our alphabet, stating that the letter at position <math>x</math> is an <math>a</math>. | |||
The semantics of the formulae in FO is straightforward, <math>\neg A</math> is true iff <math>A</math> is false, <math>A\wedge B</math> is true iff <math>A</math> is true and <math>B</math> is true, and <math>\forall x P(x) </math> is true iff <math>P(v)</math> is true for all values <math>v</math> that <math>x</math> may take in the underlying universe. For ''P'' a ''c''-ary predicate, <math>P(x_1,\dots, x_c)</math> is true if and only if when <math>x_i</math> is interpreted as <math>n_i</math> <math>P(n_1,\dots, n_c)</math> is true. | |||
== Property == | |||
=== Warning === | |||
A query in FO will then be to check if a first-order formula is true over a given structure representing the input to the problem. One should not confuse this kind of problem with checking if a quantified boolean formula is true, which is the definition of [[Quantified Boolean formula problem|QBF]], which is [[PSPACE-complete]]. The difference between those two problems is that in QBF the size of the problem is the size of the formula and elements are just boolean values, whereas in FO the size of the problem is the size of the structure and the formula is fixed. | |||
This is similar to [[Parameterized complexity]] but the size of the formula is not a fixed parameter. | |||
=== Normal form === | |||
Every formula is equivalent to a formula in [[prenex normal form]] (where all quantifiers are written first, followed a quantifier-free formula). | |||
== Operators == | |||
=== FO without any operators === | |||
In [[circuit complexity]], ''FO(ARB)'' where ''ARB'' is the set of every predicates, the logic where we can use arbitrary predicates, can be shown to be equal to [[AC0|AC<sup>0</sup>]], the first class in the [[AC (complexity)|AC]] hierarchy. Indeed, there is a natural translation from FO's symbols to nodes of circuits, with <math>\forall, \exists</math> being <math>\land</math> and <math>\lor</math> of size <math>n</math>. | |||
''FO(BIT)'' is the restriction of AC<sup>0</sup> family of circuit constructible in [[LH (complexity)|alternative logarithmic time]]. | |||
''FO(<)'' is the set of [[Star-free language]]s. | |||
=== Partial fixed point is PSPACE === | |||
FO(PFP,X) is the set of boolean queries definable in FO(X) where we add a partial fixed point operator. | |||
Let <math>k</math> be an integer, <math>x, y</math> be vectors of <math>k</math> variables, <math>P</math> be a second-order variable of arity <math>k</math>, and <math>\phi</math> be a FO(PFP,X) function using <math>x</math> and <math>P</math> as variables. We can iteratively define <math>(P_i)_{i\in N}</math> such that <math>P_0(x)=false</math> and <math>P_i(x)=\phi(P_{i-1},x)</math> (meaning <math>\phi</math> with <math>P_{i-1}</math> substituted for the second-order variable <math>P</math>). Then, either there is a fixed point, or the list of <math>(P_i)</math>s is cyclic. | |||
PFP<math>(\phi_{P,x})(y)</math> is defined as the value of the fixed point of <math>(P_i)</math> on <math>y</math> if there is a fixed point, else as false. Since <math>P</math>s are properties of arity <math>k</math>, there are at most <math>2^{n^k}</math> values for the <math>P_i</math>s, so with a polynomial-space counter we can check if there is a loop or not. | |||
It has been proven that FO(PFP,BIT) is equal to [[PSPACE]]. This definition is equivalent to FO(<math>2^{n^{O(1)}}</math>). | |||
=== Least Fixed Point is P === | |||
FO(LFP,X) is the set of boolean queries definable in FO(PFP,X) where the partial fixed point is limited to be monotone. That is, if the second order variable is <math>P</math>, then <math>P_i(x)</math> always implies <math>P_{i+1}(x)</math>. | |||
We can guarantee monotonicity by restricting the formula <math>\phi</math> to only contain positive occurrences of <math>P</math> (that is, occurrences preceded by an even number of negations). We can alternatively describe LFP(<math>\phi_{P,x}</math>) as PFP(<math>\psi_{P,x}</math>) where <math>\psi(P,x)=\phi(P,x)\vee P(x)</math>. | |||
Due to monotonicity, we only add vectors to the truth table of <math>P</math>, and since there are only <math>n^k</math> possible vectors we will always find a fixed point before <math>n^k</math> iterations. Hence it can be shown that FO(LFP,BIT)=[[P (complexity)|P]]. This definition is equivalent to FO(<math>n^{O(1)}</math>). | |||
=== Transitive closure is NL === | |||
FO(TC,X) is the set of boolean queries definable in FO(X) with a transitive closure (TC) operator. | |||
TC is defined this way: let <math>k</math> be a positive integer and <math>u,v,x,y</math> be vector of <math>k</math> variables. Then TC<math>(\phi_{u,v})(x,y)</math> is true if there exist <math>n</math> vectors of variables <math>(z_i)</math> such that <math>z_1=x, z_n=y</math>, and for all <math>i<n</math>, <math>\phi(z_i,z_{i+1})</math> is true. Here, <math>\phi</math> is a formula written in FO(TC) and <math>\phi(x,y)</math> means that the variables <math>u</math> and <math>v</math> are replaced by <math>x</math> and <math>y</math>. | |||
FO(TC,BIT) is equal to [[NL (complexity)|NL]]. | |||
=== Deterministic transitive closure is L === | |||
FO(DTC,X) is defined as FO(TC,X) where the transitive closure operator is deterministic. This means that when we apply DTC(<math>\phi_{u,v}</math>), we know that for all <math>u</math>, there exists at most one <math>v</math> such that <math>\phi(u,v)</math>. | |||
We can suppose that DTC(<math>\phi_{u,v}</math>) is [[syntactic sugar]] for TC(<math>\psi_{u,v}</math>) where <math>\psi(u,v)=\phi(u,v)\wedge \forall x, (x=v \vee \neg \psi(u,x))</math>. | |||
It has been shown that FO(DTC,BIT) is equal to [[L (complexity)|L]]. | |||
===Normal form === | |||
Any formula with a fixed point (resp. transitive cosure) operator can without loss of generality be written with exactly one application of the operators applied to 0 (resp. <math>0,(n-1)</math>) | |||
== Iterating == | |||
We will define first-order with iteration, ''''FO[<math>t(n)</math>]''''; here <math>t(n)</math> is a (class of) functions from integers to integers, and for different classes of functions <math>t(n)</math> we will obtain different complexity classes FO[<math>t(n)</math>]. | |||
In this section we will write <math>(\forall x P) Q</math> to mean <math>(\forall x (P\Rightarrow Q))</math> and <math>(\exists x P) Q</math> to mean <math>(\exists x (P \vee Q))</math>. We first need to define quantifier blocks (QB), a quantifier block is a list <math>(Q_1 x_1, phi_1)...(Q_k x_k, phi_k)</math> where the <math>phi_i</math>s are quantifier-free [[#Complexity_Zoo:F#FO|FO]]-formulae and <math>Q_i</math>s are either <math>\forall</math> or <math>\exists</math>. | |||
If <math>Q</math> is a quantifiers block then we will call <math>[Q]^{t(n)}</math> the iteration operator, which is defined as <math>Q</math> written <math>t(n)</math> time. One should pay attention that here there are <math>k*t(n)</math> quantifiers in the list, but only <math>k</math> variables and each of those variable are used <math>t(n)</math> times. | |||
We can now define FO[<math>t(n)</math>] to be the FO-formulae with an iteration operator whose exponent is in the class <math>t(n)</math>, and we obtain those equalities: | |||
*FO[<math>(\log n)^i</math>] is equal to FO-uniform [[AC (complexity)|AC<sup>i</sup>]], and in fact FO[<math>t(n)</math>] is FO-uniform AC of depth <math>t(n)</math>. | |||
*FO[<math>(\log n)^{O(1)}</math>] is equal to [[#NC_(complexity)|NC]]. | |||
*FO[<math>n^{O(1)}</math>] is equal to [[P (complexity)|PTIME]], it is also another way to write |FO(LFP). | |||
*FO[<math>2^{n^{O(1)}}</math>] is equal to [[PSPACE]], it is also another way to write [[#Complexity_Zoo:F#fopfp|FO(PFP)]]. | |||
==Logic without arithmetical relations== | |||
Let the successor relation, ''succ'', be a binary relation such that <math>\rm{succ}(x,y)</math> is true if and only if <math>x+1=y</math>. | |||
Over first order logic, ''succ'' is strictly less expressive than <, which is less expressive than +, which is less expressive than ''bit''. + and <math>\times</math> are as expressive as ''bit''. | |||
===Using successor to define ''bit'' === | |||
It is possible to define the ''plus'' and then the ''bit'' relations with a deterministic transitive closure. | |||
<math>\rm{plus}(a,b,c)=(\rm{DTC}_{v,x,y,z} \rm{succ}(v,y) \land | |||
\rm{succ}(z,x)) (a,b,c,0)</math> and | |||
<math>\rm{bit}(a,b)=(\rm{DTC}_{a,b,a',b'}\psi)(a,b,1,0)</math> with | |||
<math>\psi=\text{if } b=0 \text{ then } | |||
(\text{if } \exists m(a=m+m+1) \text{ then }(a'=1\land b'=0)\text{ else } | |||
\bot)\text{ else } (\rm{succ}(b',b) \land (a+a=a'\lor | |||
a+a+1=a')</math> | |||
This just means that when we query for bit 0 we check the parity, and go to (1,0) if <math>a</math> is odd(which is an accepting state), else we reject. If we check a bit <math>b>0</math>, we divide <math>a</math> by 2 and check bit <math>b-1</math>. | |||
Hence it makes no sense to speak of operators with successor alone, without the other predicates. | |||
===Logics without successor=== | |||
''FO[LFP]'' and ''FO[PFP]'' are two logics without any predicates, apart from the equality predicates between variables and the letters predicates. They are equal respectively to ''relational-P'' and FO(PFP) is ''relational-PSPACE'', the classes P and PSPACE over [[relational machines]].<ref name="avv">Serge Abiteboul, [[Moshe Y. Vardi]], [[Victor Vianu]]: [http://portal.acm.org/citation.cfm?id=256295|Fixpoint logics, relational machines, and computational complexity] Journal of the ACM (JACM) archive, Volume 44 , Issue 1 (January 1997), Pages: 30-56, ISSN:0004-5411</ref> | |||
The [[Abiteboul-Vianu Theorem]] states that FO(LFP)=FO(PFP) if and only if FO(<,LFP)=FO(<,PFP), hence if and only if P=PSPACE. This result has been extended to other fixpoints.<ref name="avv"/> This shows that the order problem in first order is more a technical problem than a fundamental one. | |||
== References == | |||
* [[Ronald Fagin]], [http://www.almaden.ibm.com/cs/people/fagin/genspec.pdf Generalized First-Order Spectra and Polynomial-Time Recognizable Sets]. ''Complexity of Computation'', ed. R. Karp, SIAM-AMS Proceedings 7, pp. 27–41. 1974. | |||
* Ronald Fagin, [http://www.almaden.ibm.com/cs/people/fagin/tcs93.pdf Finite model theory-a personal perspective]. Theoretical Computer Science 116, 1993, pp. 3–31. | |||
* Neil Immerman. [http://www.cs.umass.edu/~immerman/pub/capture.pdf Languages Which Capture Complexity Classes]. ''15th ACM STOC Symposium'', pp. 347–354. 1983. | |||
{{Reflist}} | |||
== External links == | |||
* [http://www.cs.umass.edu/~immerman/descriptive_complexity.html Neil Immerman's descriptive complexity page], including a diagram | |||
* [http://qwiki.stanford.edu/wiki/Complexity_Zoo:F#fo|Complexity zoo about FO], see the class under it also | |||
{{DEFAULTSORT:Fo (Complexity)}} | |||
[[Category:Descriptive complexity| ]] | |||
[[Category:Finite model theory]] | |||
[[Category:Complexity classes]] |
Revision as of 07:55, 29 December 2013
In descriptive complexity, a branch of Computational complexity FO is a complexity class of structures which can be recognized by formulas of first-order logic. It is the foundation of the field of descriptive complexity and is equal to the complexity class AC0. Descriptive complexity uses First-order logic formalism, but does not use most of the usual notion associated to logic as proof theory or axiomatization.
When the predicates are restricted to a set X, FO[X] correspond to some other class, and there exists specific sets X such FO[X] has some interesting properties. In particular, FO[<] is equal to the set of Star-free languages. Having two different definition, in term of logic and in term of regular expression, is a hint showing that this class may have some mathematical interest, and that we may use together methods for both domain to do proofs.
Similarly, various extensions of FO, formed by the addition of certain operators, give rise to other well-known complexity classes,[1] allowing the complexity of some problems to be proven without having to look at them as algorithm problem.
Definition and examples
The idea
When we use the logic formalism to describe a computational problem, the input is a finite structure, and the elements of that structure are the domain of discourse. Usually the input is either a string (of bits or over an alphabet) the elements of which are positions of the string, or a graph of which the elements are vertices. The length of the input will be measured by the size of the respective structure. Whatever the structure is, we can assume that there are relations that can be tested, for example " is true iff there is an edge from to " (in case of the structure being a graph), or " is true iff the th letter of the string is 1." These relations are the predicates for the first-order logic system. We also have constants, which are special elements of the respective structure, for example if we want to check reachability in a graph, we will have to choose two constants s (start) and t (terminal).
In descriptive complexity theory we almost always suppose that there is a total order over the elements and that we can check equality between elements. This lets us consider elements as numbers: the element represents the number iff there are elements with . Thanks to this we also may have the primitive predicate "bit", where is true if only the th bit of is 1. (We can replace addition and multiplication by ternary relations such that is true iff and is true iff ).
Formally
Let X be a set of predicate. The language FO[X] is defined as the closure by conjunction ( ), negation () and universal quantification () over elements of the structures. We also often use existential quantification () and disjunction () but those can be defined by means of the first 3 symbols. The base case being the predicates of X applied to some variables. We always implicitely have a predicate for each letter of our alphabet, stating that the letter at position is an .
The semantics of the formulae in FO is straightforward, is true iff is false, is true iff is true and is true, and is true iff is true for all values that may take in the underlying universe. For P a c-ary predicate, is true if and only if when is interpreted as is true.
Property
Warning
A query in FO will then be to check if a first-order formula is true over a given structure representing the input to the problem. One should not confuse this kind of problem with checking if a quantified boolean formula is true, which is the definition of QBF, which is PSPACE-complete. The difference between those two problems is that in QBF the size of the problem is the size of the formula and elements are just boolean values, whereas in FO the size of the problem is the size of the structure and the formula is fixed.
This is similar to Parameterized complexity but the size of the formula is not a fixed parameter.
Normal form
Every formula is equivalent to a formula in prenex normal form (where all quantifiers are written first, followed a quantifier-free formula).
Operators
FO without any operators
In circuit complexity, FO(ARB) where ARB is the set of every predicates, the logic where we can use arbitrary predicates, can be shown to be equal to AC0, the first class in the AC hierarchy. Indeed, there is a natural translation from FO's symbols to nodes of circuits, with being and of size .
FO(BIT) is the restriction of AC0 family of circuit constructible in alternative logarithmic time. FO(<) is the set of Star-free languages.
Partial fixed point is PSPACE
FO(PFP,X) is the set of boolean queries definable in FO(X) where we add a partial fixed point operator.
Let be an integer, be vectors of variables, be a second-order variable of arity , and be a FO(PFP,X) function using and as variables. We can iteratively define such that and (meaning with substituted for the second-order variable ). Then, either there is a fixed point, or the list of s is cyclic.
PFP is defined as the value of the fixed point of on if there is a fixed point, else as false. Since s are properties of arity , there are at most values for the s, so with a polynomial-space counter we can check if there is a loop or not.
It has been proven that FO(PFP,BIT) is equal to PSPACE. This definition is equivalent to FO().
Least Fixed Point is P
FO(LFP,X) is the set of boolean queries definable in FO(PFP,X) where the partial fixed point is limited to be monotone. That is, if the second order variable is , then always implies .
We can guarantee monotonicity by restricting the formula to only contain positive occurrences of (that is, occurrences preceded by an even number of negations). We can alternatively describe LFP() as PFP() where .
Due to monotonicity, we only add vectors to the truth table of , and since there are only possible vectors we will always find a fixed point before iterations. Hence it can be shown that FO(LFP,BIT)=P. This definition is equivalent to FO().
Transitive closure is NL
FO(TC,X) is the set of boolean queries definable in FO(X) with a transitive closure (TC) operator.
TC is defined this way: let be a positive integer and be vector of variables. Then TC is true if there exist vectors of variables such that , and for all , is true. Here, is a formula written in FO(TC) and means that the variables and are replaced by and .
FO(TC,BIT) is equal to NL.
Deterministic transitive closure is L
FO(DTC,X) is defined as FO(TC,X) where the transitive closure operator is deterministic. This means that when we apply DTC(), we know that for all , there exists at most one such that .
We can suppose that DTC() is syntactic sugar for TC() where .
It has been shown that FO(DTC,BIT) is equal to L.
Normal form
Any formula with a fixed point (resp. transitive cosure) operator can without loss of generality be written with exactly one application of the operators applied to 0 (resp. )
Iterating
We will define first-order with iteration, 'FO[]'; here is a (class of) functions from integers to integers, and for different classes of functions we will obtain different complexity classes FO[].
In this section we will write to mean and to mean . We first need to define quantifier blocks (QB), a quantifier block is a list where the s are quantifier-free FO-formulae and s are either or . If is a quantifiers block then we will call the iteration operator, which is defined as written time. One should pay attention that here there are quantifiers in the list, but only variables and each of those variable are used times.
We can now define FO[] to be the FO-formulae with an iteration operator whose exponent is in the class , and we obtain those equalities:
- FO[] is equal to FO-uniform ACi, and in fact FO[] is FO-uniform AC of depth .
- FO[] is equal to NC.
- FO[] is equal to PTIME, it is also another way to write |FO(LFP).
- FO[] is equal to PSPACE, it is also another way to write FO(PFP).
Logic without arithmetical relations
Let the successor relation, succ, be a binary relation such that is true if and only if .
Over first order logic, succ is strictly less expressive than <, which is less expressive than +, which is less expressive than bit. + and are as expressive as bit.
Using successor to define bit
It is possible to define the plus and then the bit relations with a deterministic transitive closure.
This just means that when we query for bit 0 we check the parity, and go to (1,0) if is odd(which is an accepting state), else we reject. If we check a bit , we divide by 2 and check bit .
Hence it makes no sense to speak of operators with successor alone, without the other predicates.
Logics without successor
FO[LFP] and FO[PFP] are two logics without any predicates, apart from the equality predicates between variables and the letters predicates. They are equal respectively to relational-P and FO(PFP) is relational-PSPACE, the classes P and PSPACE over relational machines.[2]
The Abiteboul-Vianu Theorem states that FO(LFP)=FO(PFP) if and only if FO(<,LFP)=FO(<,PFP), hence if and only if P=PSPACE. This result has been extended to other fixpoints.[2] This shows that the order problem in first order is more a technical problem than a fundamental one.
References
- Ronald Fagin, Generalized First-Order Spectra and Polynomial-Time Recognizable Sets. Complexity of Computation, ed. R. Karp, SIAM-AMS Proceedings 7, pp. 27–41. 1974.
- Ronald Fagin, Finite model theory-a personal perspective. Theoretical Computer Science 116, 1993, pp. 3–31.
- Neil Immerman. Languages Which Capture Complexity Classes. 15th ACM STOC Symposium, pp. 347–354. 1983.
43 year old Petroleum Engineer Harry from Deep River, usually spends time with hobbies and interests like renting movies, property developers in singapore new condominium and vehicle racing. Constantly enjoys going to destinations like Camino Real de Tierra Adentro.
External links
- Neil Immerman's descriptive complexity page, including a diagram
- zoo about FO, see the class under it also
- ↑ N. Immerman Descriptive complexity (1999 Springer)
- ↑ 2.0 2.1 Serge Abiteboul, Moshe Y. Vardi, Victor Vianu: logics, relational machines, and computational complexity Journal of the ACM (JACM) archive, Volume 44 , Issue 1 (January 1997), Pages: 30-56, ISSN:0004-5411