Noncommutative Gauge Theories from Deformation Quantization

合集下载

杨振宁_二十世纪物理学的三个主旋律

杨振宁_二十世纪物理学的三个主旋律
20
Heisenberg’s mechanics has restored my zest for life.
Pauli to Kronig, October 9, 1925
21
Wolftgang Pauli (1900-1958) 22
Do not enter into this conflict, we are both much too kind and gentle to participate in that kind of struggle. Both Bohr and Heisenberg are tough, hard nosed, uncompromising and
31
Erwin Schrödinger (1887-1961) 32
Enrico Fermi (1901-1954)
33
Pauli
— 强力
Fermi
— 坚实
Heisenberg — 深入见解ຫໍສະໝຸດ Dirac— 纯真
34
2. Symmetry (= invariance)
35
The five regular solids with maximum symmetry. Reprinted
Quantization Symmetry
Phase Factor
10
1. Quantization
11
Max Planck (1858-1947)
12
13
14
Quantization
1900 1905 1913
Planck Einstein Bohr
15
Albert Einstein (1879-1955)
It was a period of patient work in the laboratory, of crucial experiments and daring action, of many false starts and many untenable conjectures. It was a time of earnest correspondence and

Theories with global gauge anomalies on the lattice

Theories with global gauge anomalies on the lattice

a r X i v :h e p -l a t /0010003v 1 3 O c t 20001Theories with global gauge anomalies on the latticeP.Mitra a∗aSaha Institute of Nuclear Physics,Block AF,Bidhannagar,Calcutta 700064,INDIAA global anomaly in a chiral gauge theory manifests itself in different ways in the continuum and on the lattice.In the continuum case,functional integration of the fermion determinant over the whole space of gauge fields yields zero.In the case of the lattice,it is not even possible to define a fermion measure over the whole space of gauge configurations.However,this is not necessary,and as in the continuum,a reduced functional integral is sufficient for the existence of the theory.Presented at Lattice 2000,Bangalorehep-lat/00100031.IntroductionAnomalies are of two different types.Local or divergence anomalies have been known since 1969[1]:classically conserved symmetry currents cease to be conserved after quantization if there are anomalies of this kind.For example,the the-oryL =¯ψ(i∂/−eA /)ψ−116π2F µνF µν=0.(2)If an anomalous current is associated with a gauged symmetry ,it leads to an apparent problem in quantization because the equations of motion of the gauge fields require the current to be con-served.A treatment of such a theory just like a usual gauge theory shows an inconsistency.This problem can be sorted out by paying proper at-tention to phase space constraints,as suggested by [2].The anomaly itself can be made to vanish in a sense by going to the constrained subman-ifold of classical phase space.However,theories with anomalous gauge currents are to be distin-guished from theories with nonanomalous gauge currents.If a theory is nonanomalous ,it pos-sesses gauge freedom,and is describable in any22.Functional integral in the continuum The full partition function of the gauge theory with fermions may be written asZ= D AZ[A],(3) Z[A]≡e−S eff= D¯ψDψe−S(ψ,¯ψ,A)(4)An anomaly-free theory has Z[A]gauge invariant. If there is a gauge anomaly,Z[A]varies under gauge transformations of A:Z[A g]=e iα(A,g−1)Z[A],(5) whereαmay be regarded as an integral represen-tation of the anomaly.It obeys some consistency conditions(mod2π):α(A,g−12g−11)=α(A,g−11)+α(A g1,g−12)α(A,g−1)=−α(A g,g).(6) The case becomes one of a global anomaly ifαis independent of A,and vanishes for g connected to the identity but not for some g which cannot be continuously connected to the identity.A-independence implies an abelian representation satisfyingα(g2g1)=α(g1)+α(g2).(7) In the SU(2)case,the two components of the gauge group manifest themselves in two possible values of the phase:e iα=±1.In an anomaly-free theory,the partition func-tion factorizes into the volume of the gauge group and the gauge-fixed partition function:Z= D AZ[A]= D AZ[A] D gδ(f(A g))∆f(A)= D g D AZ[A g−1]δ(f(A))∆f(A) = D g D AZ[A]δ(f(A))∆f(A)=( D g)Z f(8)This is the standard Faddeev-Popov argument. Here,δ(f)represents a gauge-fixing operation and∆f is the corresponding Faddeev-Popov de-terminant.This decoupling of gauge degrees of freedom does not occur if a local anomaly is present.For a global anomaly however,the partition function factorizes again:Z= D ge−iα(g) D AZ[A]δ(f(A))∆f(A)(9)As the phase factors form a representation of the gauge group,D ge−iα(g)= D(gh)e−iα(gh)=e−iα(h) D ge−iα(g)(10)where h stands for afixed gauge transformation. If h is not connected to the identity,e−iα(h)=1, and consequently D ge−iα(g)=0,which in turn means that Z=0.Does this mean that the the-ory cannot be defined?Let us look at expectation values of gauge invariant operators.D AZ[A]OD ge−iα(g) D AZ[A]δ(f(A))∆f(A)(11) The expression on the right is of the form0D AZ[A]δ(f(A))∆f(A).(12)The right hand side is precisely what one gets in the canonical approach to quantization where gauge degrees of freedom are removed byfixing the gauge at the classical level and only physi-cal degrees of freedom enter the functional inte-gral.The Faddeev-Popov determinant arises in the canonical approach as the determinant of the matrix of Poisson brackets of what may be called the”second class constraints”,i.e.,the Gauss law3operator and the gaugefixing condition f,which is of course introduced by hand and not really a constraint of the theory.There are both ordinary fields and conjugate momenta,but the latter are easily integrated over.The point is that the full functional integral is not needed in the canonical approach and there is no harm if it vanishes!A trace is left behind by the global anomaly. One may imagine a classification of the gauge-fixing functions f where f,f′are said to belong to the same class if there exists a gauge transfor-mation connected to the identity to go from a con-figuration with f=0to one with f′=0.Then Z f=Z f′.More generally,when such a transfor-mation is not connected to the identity,Z f=e−iα(g0)Z f′,(13) where g0is determined by f,f′.These factors e−iα(g0)occurring in partition functions cancel out in expectation values of gauge invariant oper-ators,so that Green functions of gauge invariant operators are fully gauge independent[4]. There is an assumption in all this:that there is a possibility offixing the gauge.A general the-orem[6]asserts that gauges cannot befixed in a smooth way.For the construction of functional integrals,however,it is sufficient to have piece-wise smooth gauges.It should also be remem-bered that these questions arise even for theories without disconnected gauge groups and are not specific to the context of global anomalies.ttice formulationOn going to the lattice,one starts to use group-valued variables associated with links instead of A defined at points of the continuum.The topol-ogy also changes:the gauge group becomes con-nected on the lattice:it becomes possible to go to any gauge transformation from the identity in a continuous manner.Thus there are no large gauge transformations any more.Does it mean that there is no global anomaly on the lattice? The issue is complicated because chiral symme-try is not straightforward here.Chiral symmetry on the lattice has begun to make more sense in the last few years thanks to the Ginsparg-Wilson relation imposed on D,the euclidean lattice Dirac operator:γ5D+Dγ5=aDγ5D,(14) where a is the lattice spacing.An analogue ofγ5 appears from the above relation:γ5D=−DΓ5,Γ5≡γ5(1−aD).(15) It satisfies(Γ5)2=1,(Γ5)†=Γ5,(16) and can be used to define left-handed projection: P−ψ≡12[1+γ5]=¯ψ.(17) In this way of defining chiral projections,P−,but not P+,depends on the gaugefield configuration. Nontriviality of chirality on the lattice stems from this P−.A fermion measure is defined by specifying a basis of lattice Diracfields v j(x)satisfyingP−v j=v j,(v j,v k)=δjk.(18) One has to integrate over Grassmann-valued ex-pansion coefficients inψ(x)= j a j v j(x).(19)Expansion coefficients also come from the expan-sion of¯ψin terms of¯v j satisfying¯v j P+=¯v j, but these are as usual,i.e.,do not involve gauge fields.Questions of locality and integrability arise be-cause of the gaugefield dependence in P−.Ab-sence of a local anomaly appears to be sufficient to ensure locality[7].Global anomalies are man-ifested as a lack of integrability.Consider,following[5],a closed path in the SU(2)gauge configuration space,with the param-eter t running from0to1.Definef(t)=det[1−P++P+D(t)Q t D(0)†],(20) with D(t)the Dirac operator corresponding to gaugefields at parameter value t,and Q t the uni-tary transport operator for P−(t)defined by∂t Q t=[∂t P−(t),P−(t)]Q t,Q0=1.(21)4Then f(t)is real,positive and satisfiesf(1)=T f(0).(22)HereT=det[1−P−(0)+P−(0)Q1]=±1(23)depending on the topology of the considered path in the gauge configuration space.f changes sign an even or odd number of times along path de-pending on T and while det D(t)is related to f2, det D(t)det D(0)†=f2(t),(24)the chiral fermion determinant det Dχ(t)behaves like f:det Dχ(t)det Dχ(0)†=f(t)W(t)−1, (Dχ)ij≡a4 x¯v i(x)Dv j(x).(25) Here W(t)is a phase factor arising from the gauge field dependence of v j.It is a lattice artifact and may be taken to reduce to unity near the contin-uum limit.Then det Dχchanges sign,i.e.,fails to return to its starting value after transportation along a closed path if the path hasT=−1.(26)Such paths have been shown to exist in the SU(2) theory.A part of such a path lies along a gauge orbit,and a part is non-gauge.2Thus det Dχis multivalued,implying that the fermion measure is not well defined,and hence the functional integral does not make sense.This is roughly similar to the continuum.The Dirac operator is gauge-invariant and its determinant and f can change only on non-gauge portions of the closed path.So the problem of sign change of f occurs once again in non-gauge paths connect-ing gauge-related configurations.However,in the continuum,the sign change occurs between con-figurations which can be connected only by a non-gauge path.On the lattice,the sign change oc-curs when configurations are connected by a non-gauge path,though a connection is also possible。

哥德尔不完备定理英文原文

哥德尔不完备定理英文原文

哥德尔不完备定理英文原文英文回答:Gödel's incompleteness theorems are two mathematical theorems that demonstrate inherent limitations of axiomatic systems based on first-order logic. The theorems were published by Kurt Gödel in 1931 and are widely acknowledged as foundational results in mathematical logic.Theorem 1 (Incompleteness theorem): Any effectively axiomatizable theory capable of expressing basic arithmetic is either incomplete or inconsistent. That is, there are true statements about the natural numbers that cannot be proven within the theory.Theorem 2 (Undecidability theorem): No consistent, effectively axiomatizable theory capable of expressing basic arithmetic can decide all true statements about the natural numbers. That is, there are statements about the natural numbers that can neither be proven nor disprovenwithin the theory.Implications of Gödel's theorems:Limits of formal systems: Gödel's theorems demonstrate that no formal system can be both complete and consistent if it is capable of expressing basic arithmetic. This has profound implications for the foundations of mathematics and the limits of what can be proven within a given axiomatic system.Creativity and human intelligence: The incompleteness theorems suggest that there are mathematical truths that cannot be discovered through purely mechanical or algorithmic processes. This has led to speculation that human intelligence may involve non-computational elements that allow for creativity and insight.The nature of mathematics: Gödel's theorems have led to a deeper understanding of the nature of mathematics. They have helped to establish the distinction between provability and truth, and have raised questions about therole of intuition and human understanding in mathematical reasoning.中文回答:哥德尔不完备定理。

Quantum Groups and Noncommutative Geometry

Quantum Groups and Noncommutative Geometry

a r X i v :h e p -t h /0006167v 1 21 J u n 2000QUANTUM GROUPS AND NONCOMMUTATIVE GEOMETRYShahn MajidSchool of Mathematical Sciences,Queen Mary and Westfield College University of London,Mile End Rd,London E14NS,UK November,1999Abstract Quantum groups emerged in the latter quarter of the 20th century as,on the one hand,a deep and natural generalisation of symmetry groups for certain integrable systems,and on the other as part of a generalisation of geometry itself powerful enough to make sense in the quantum domain.Just as the last century saw the birth of classical geometry,so the present century sees at its end the birth of this quantum or noncommutative geometry,both as an elegant mathematical reality and in the form of the first theoretical predictions for Planck-scale physics via ongoing astronomical measurements.Noncommutativity of spacetime,in particular,amounts to a postulated new force or physical effect called cogravity.I Introduction Now that quantum groups and their associated quantum geometry have been around for more than a decade,it is surely time to take stock.Where did quantum groups come from,what have they achieved and where are they going?This article,which is addressed to non-specialists (but should also be interesting for experts)tries to answer this on two levels.First of all on the level of quantum groups themselves as mathematical tools and building blocks for physical models.And,equally importantly,quantum groups and their associated noncommutative geometry in terms of their overall significance for mathematics and theoretical physics,i.e.,at a more conceptual level.Obviously this latter aspect will be very much my own perspective,which is that of a theoretical physicist who came to quantum groups a decade ago as a tool to unify quantum theory and gravity in an algebraic approach to Planck scale physics.This is in fact only one of the two main origins in physics of quantum groups;the other being integrable systems,which I will try to cover as well.Let me also say that noncommutative geometry has other approaches,notably the one of A.Connes coming out of operator theory.I will say something about this too,although,until recently,this has largely been a somewhat different approach.We start with the conceptual significance for theoretical physics.It seems clear to me that future generations looking back on the 20th century will regard the discovery of quantum mechanics in the 1920s,i.e.the idea to replace the coordinates x,p of classical mechanics bynoncommuting operators x,p,as one of its greatest achievements in our understanding of Nature, matched in its significance only by the unification of space and time as a theory of gravity.But whereas the latter was well-founded in the classical geometry of Newton,Gauss,Riemann and Poincar´e,quantum theory was something much more radical and mysterious.Exactly which variables in the classical theory should correspond to operators?They are local coordinates on phase space but how does the global geometry of the classical theory look in the quantum theory,what does it fully correspond to?The problem for most of this century was that the required mathematical structures to which the classical geometry might correspond had not been invented and such questions could not be answered.As I hope to convince the reader,quantum groups and their associated noncommutative geometry have led in the last decades of the20th century to thefirst definitive answers to this kind of question.There has in fact emerged a more or less systematic generalisation of geometry every bit as radical as the step from Euclidean to non-Euclidean,and powerful enough not to break down in the quantum domain.I do doubt very much that what we know today will be thefinal formulation,but it is a definitive step in a right and necessary direction and a turning point in the future development of mathematical and theoretical physics.For example,any attempt to build a theory of quantum gravity with classical starting point a smooth manifold–this includes loop-variable quantum gravity,string theory and quantum cosmology,is necessarily misguided except as some kind of effective approximation:smooth manifolds should come out of the algebraic structure of the quantum theory and not be a starting point for the latter. There is no evidence that the real world is any kind of smooth continuum manifold except as a macroscopic approximation and every reason to think that it is fundamentally not.I therefore doubt that any one of the above could be a‘theory everything’until it becomes an entirely algebraic theory founded in noncommutative geometry of some kind or other.Of course,this is my personal view.At any rate,I do not think that the fundamental importance of noncommutative geometry can be overestimated.First of all,anyone who does quantum theory is doing noncommutative geometry whether wanting to admit it or not,namely noncommutative geometry of the phase space.Less obvious but also true,we will see in Section II that if the position space is curved then the momentum space is by itself intrinsically noncommutative.If one gets this far then it is also natural that the position space or spacetime by itself could be noncommutative,which would correspond to a curved or nonAbelian momentum group.This is one of the bolder predictions coming out of noncommutative geometry.It has the simple physical interpretation as what I call cogravity,i.e.curvature or‘gravity’in momentum space.As such it is independent of i.e. dual to curvature or gravity in spacetime and would appear as a quite different and new physical effect.Theoretically cogravity can,for example,be detected as energy-dependence of the speed8185909510000500981200yearpapersFigure 1:Growth of research papers on quantum groupsof light.Moreover,even if cogravity was very weak,of the order of a Planck-scale effect,it could still in principle be detected by astronomical measurements at a cosmological level.Therefore,just in time for the new millennium,we have the possibility of an entirely new physical effect in Nature coming from fresh and conceptually sound new mathematics .Where quantum groups precisely come into this is as follows.Just as Lie groups and their associated homogeneous spaces provided definitive examples of classical differential geometry even before Riemann formulated their intrinsic structure as a theory of manifolds,so quantum groups and their associated quantum homogeneous spaces,quantum planes etc.,provide large (i.e.infinite)classes of examples of proven mathematical and physical worth and clear geomet-rical content on which to build and develop noncommutative differential geometry.They are noncommutative spaces in the sense that they have generators or ‘coordinates’like the non-commuting operators x ,p in quantum mechanics but with a much richer and more geometric algebraic structure than the Heisenberg or CCR algebra.In particular,I do not believe that one can build a theory of noncommutative differential geometry based on only one example such as the Heisenberg algebra or its variants (however fascinating)such as the much-studied noncommutative torus.One needs many more ‘sample points’in the form of natural and varied examples to obtain a valid general theory.By contrast,if one does a search of BIDS one finds,see Figure 1,vast numbers of papers in which the rich structure and applications of quantum groups are explored and justified in their own right (data complied from BIDS:published pa-pers since 1981with title or abstract containing ‘quantum group*’,‘Hopf alg*’,‘noncommutative geom*’,‘braided categ*’,‘braided group*’,‘braided Hopf*’.)This is the significance of quantum groups.And of course something like them should be needed in a quantum world where there is no evidence for a classical space such as the underlying set of a Lie group.Finally,it turns out that noncommutative geometry,at least of the type that we shall de-scribe,is in many ways cleaner and more straightforward than the special commutative limit. One simply does not need to assume commutativity in most geometrical constructions,including differential calculus and gauge theory.The noncommutative version is often less infinite,dif-ferentials are often more regularfinite-differences,etc.And noncommutative geometry(unlike classical geometry)can be specialised without effort to discrete spaces or tofinite-dimensional algebras.It is simply a powerful and natural generalisation of geometry as we usually know it. So my overall summary and prediction for the next millennium from this point of view is:•All geometry will be noncommutative(or whatever comes beyond that),with conventional geometry merely a special case.•The discovery of quantum theory,its correspondence principle(and noncommutative ge-ometry is nothing more than the elaboration of that)will be considered one of the century’s greatest achievement in mathematical physics,commensurate with the discovery of clas-sical geometry by Newton some centuries before.•Quantum groups will be viewed as thefirst nontrivial class of examples and thereby point-ers to the correct structure of this noncommutative geometry.•Spacetime too(not only phase space)will be known to be noncommutative(cogravity will have been detected).•At some point a future Einstein will combine the then-standard noncommutative geomet-rical ideas with some deep philosophical ideas and explain something really fundamental about our physical reality.In the fun spirit of this article,I will not be above putting down my own thoughts on this last point.These have to do with what I have called for the last decade the Principle of representation-theoretic self-duality[1].In effect,it amounts to extending the ideas of Born reciprocity,Mach’s principle and Fourier theory to the quantum domain.Roughly speaking, quantum gravity should be recast as gravity and cogravity both present and dual to each other and with Einstein’s equation appearing as a self-duality condition.The longer-term philosophical implications are a Kantian or Hegelian view of the nature of physical reality,which I propose in Section V as a new foundation for next millennium.We now turn to another fundamental side of quantum groups,which is at the heart of their other origin in physics,namely as generalised symmetry groups in exactly solvable lattice models. It leads to diverse applications ranging from knot theory to representation theory to Poisson geometry,all areas that quantum groups have revolutionised.What is really going on here in my opinion is not so much the noncommutative geometry of quantum groups themselves as a different kind of noncommutativity or braid statistics which certain quantum groups induce onany objects of which they are a symmetry.The latter is what I have called‘noncommutativity of the second kind’or outer noncommutativity since it not so much a noncommutativity of one algebra as a noncommutative modification of the exchange law or tensor product of any two independent algebras or systems.It is the notion of independence which is really being deformed here.Recall that the other great‘isation’idea in mathematical physics in this century(after ‘quantisation’)was‘superisation’,where everything is Z2-graded and this grading enters into how two independent systems are interchanged.Physics traditionally has a division into bosonic or force particles and fermionic or matter particles according to this grading and exchange behaviour.So certain quantum groups lead to a generalisation of that as braided geometry[2]or a process of braidification.These quantum groups typically have a parameter q and its meaning is a generalisation of the−1for supersymmetry.This in turn leads to a profound generalisation of conventional(including super)mathematics in the form of a new concept of algebra wherin one‘wires up’algebraic operations much as the wiring in a computer,i.e.outputs of one into inputs of another.Only,this time,the under or over crossings are nontrivial(and generally distinct)operations depending on q.These are the so-called‘R-matrices’.Afterwards one has the luxury of both viewing q in this way or expanding it around1in terms of a multiple of Planck’s constant and calling it a formal‘quantisation’–q-deformation actually unifies both ‘isation’processes.For example,Lorentz-invariance,by the time it is q-deformed[3],induces braid statistics even when particles are initially bosonic.In summary,•The notion of symmetry or automorphism group is an artifact of classical geometry and ina quantum world should naturally be generalised to something more like a quantum groupsymmetry.•Quantum symmetry groups induce braid statistics on the systems on which they act.In particular,the notion of bose-fermi statistics or the division into force and matter particles is an artifact of classical geometry.•Quantisation and the departure from bosonic statistics are two limits of the same phe-nomenon of braided geometry.Again,there are plenty of concrete models in solid state physics already known with quantum group symmetry.The symmetry is useful and can be viewed(albeit with hindsight)as the origin of the exact solvability of these models.These two points of view,the noncommutative geometrical and the generalised symmetry, are to date the two main sources of quantum groups.One has correspondingly two mainflavours or types of quantum groups which really allowed the theory to take off.Both were introduced at the mid1980s although the latter have been more extensively studied in terms of applicationsto date.They include the deformationsU q(g)(1)of the enveloping algebra U(g)of every complex semisimple Lie algebra g[4][5].These have as many generators as the usual ones of the Lie algebra but modified relations and,additionally, a structure called the‘coproduct’.The general class here is that of quasitriangular quantum groups.They arose as generalised symmetries in certain lattice models but are also visible in the continuum limit quantumfield theories(such as the Wess-Zumino-Novikov-Witten model on the Lie group G with Lie algebra g).The coordinate algebras of these quantum groups are further quantum groups C q[G]deforming the commutative algebra of coordinate functions on G.There is again a coproduct,this time expressing the group law or matrix multiplication.Meanwhile, the type coming out of Planck scale physics[6]are the bicrossproduct quantum groupsC[M]◮⊳U(g)(2)associated to the factorisation of a Lie group X into Lie subgroups,X=GM.Here the in-gredients are the conventional enveloping algebra U(g)and the commutative coordinate algebra C[M].The factorisation is encoded in an action and coaction of one on the other to make a semidirect product and coproduct◮⊳.These quantum arose at about the same time but quite independently of the U q(g),as the quantum algebras of observables of certain quantum spaces. Namely it turns out that G acts on the set M(and vice-versa)and the quantisation of those orbits are these quantum groups.This means that they are literally noncommutative phase spaces of honest quantum systems.In particular,every complex semisimple g has an associated complexification and its Lie group factorises G C=GG⋆(the classical Iwasawa decomposition) so there is an exampleC[G⋆]◮⊳U(g)(3)built from just the same data as for U q(g).In fact the Iwasawa decomposition can be understood in Poisson-Lie terms with g⋆the classical‘Yang-Baxter dual’of g.In spite of this,there is,even after a decade of development,no direct connection between the two quantum groups:gւց(4)U q(g)←?→C[G⋆]◮⊳U(g).They are both‘exponentiations’of the same classical data but apparently of completely different type(this remains a mystery to date.)Figure2:The landscape of noncommutative geometry todayAssociated to these twoflavours of quantum groups there are corresponding homogeneous spaces such as quantum spheres,quantum spacetimes,etc.Thus,of thefirst type there is a q-Minkowski space introduced in[7]as a q-Lorentz covariant algebra,and independently about a year later in[8]as2×2braided hermitian matrices.It is characterised by[x i,t]=0,[x i,x j]=0.(5) Meanwhile,of the second type there is a noncommutativeλ-Minkowski space with[x i,t]=λx i,[x i,x j]=0(6)which is the one that provides thefirst known predictions testable by astronomical measurements (by gamma-ray bursts of cosmological origin[9]).This kind of algebra was proposed as spacetime in[10]and in the4-dimensional case it was shown in[11]to be covariant under a Poincar´e quantum group of bicrossproduct form.These are clearly in sharp contrast.There are of course many more objects than these.q-spheres,q-planes etc.In Section IV we turn to the notion of‘quantum manifold’that is emerging from all these examples.Riemann was able to formulate the notion of Riemannian manifold as a way to capture known examples like spheres and tori but broad enough to formulate general equations for the intrinsic structure of space itself(or after Einstein,space-time).We are at a similar point now and what this ‘quantum groups approach to noncommutative geometry’is is more or less taking shape.It has the same degree of‘flabbiness’as Riemannian geometry(it is not tied to specific integrable systems etc.)while at the same time it includes the‘zoo’of already known naturally occurring examples,mostly linked to quantum groups.Such things as Ricci tensor and Einstein’s equation are not yet understood from this approach,however,so I would not say it is the last word.This approach is in fairly sharp contrast to‘traditional’noncommutative geometry as it was done before the emergence of quantum groups.That theory was developed by mathematiciansand mathematical physicists also coming from quantum mechanics but being concerned more with topological completions and Hilbert spaces.Certainly a beautiful theory of von-Neumann and C∗algebras emerged as an analogue of point-set topology.Some general methods such as cyclic cohomology were also developed in the1970s,with remarkable applications throughout mathematics[12].However,for concrete examples with actual noncommutative differential geometry one usually turned either to an actual manifold as input datum or to the Weyl algebra (or noncommutative torus)defined by relationsvu=e2πıθuv.(7)This in turn is basically the usual CCR or Heisenberg algebra[x,p]=ı (8)in exponentiated form.And at an algebraic level(i.e.until one considers the precise C∗-algebra completion)this is basically the usual algebra B(H)of operators on a Hilbert space as in quantum mechanics.Or at roots of unity it is M n(C)the algebra of n×n matrices.So at some level these are all basically one example.Unfortunately many of the tricks one can pull for this kind of example are special to it and not a foundation for noncommutative differential geometry of the type we need.For example,to do gauge theory Connes and M.Rieffel[13] used derivations for two independent vectorfields on the torus.The formulation of‘vector field’as a derivation of the coordinate algebra is what I would call the traditional approach to noncommutative geometry.For quantum groups such as C q[G]one simply does not have those derivations(rather,they are in general braided derivations).Similarly,in the traditional approach one defines a‘vector bundle’as afinitely-generated projective module without any of the infrastructure of differential geometry such as a principal bundle to which the vector bundle might be associated,etc.All of that could not emerge until quantum groups arrived(one clearly should take a quantum group asfiber).This is how the quantum groups approach differs from the work of Connes,Rieffel,Madore and others.It is also worth noting that string theorists have recently woken up to the need for a noncommutative spacetime but,so far at least,have still considered only this‘traditional’Heisenberg-type algebra.In the last year or two there has been some success in merging these approaches,however;a trend surely to be continued. By now both approaches have a notion of‘noncommutative manifold’which appear somewhat different but which have as point of contact the Dirac operator.Preliminaries.A full text on quantum groups is[14].To be self-contained we provide here a quick defiter on we will see many examples and various justifications for this concept. Thus,a quantum group or Hopf algebra is•A unital algebra H,1over thefield C(say)•A coproduct∆:H→H⊗H and counitǫ:H→C forming a coalgebra,with∆,ǫalgebra homomorphisms.•An antipode S:H→H such that·(S⊗id)∆=1ǫ=·(id⊗S)∆.Here a coalgebra is just like an algebra but with the axioms written as maps and arrows on the maps reversed.Thus the coassociativity and counity axioms are(∆⊗id)∆=(id⊗∆)∆,(ǫ⊗id)∆=(id⊗ǫ)∆=id.(9)The antipode plays a role that generalises the concept of group inversion.Other than that the only new mathematical structure that the reader has to contend with is the coproduct∆and its associated counit.There are several ways of thinking about the meaning of this depending on our point of view.If the quantum group is like the enveloping algebra U(g)generated by a Lie algebra g,one should think of∆as providing the rule by which actions extend to tensor products.Thus,U(g)is trivially a Hopf algebra with∆ξ=ξ⊗1+1⊗ξ,∀ξ∈g,(10)which says that when a Lie algebra elementξacts on tensor products it does so byξin the first factor and thenξin the second factor.Similarly it says that when a Lie algebra acts on an algebra it does so as a derivation.On the other hand,if the quantum group is like a coordinate algebra C[G]then∆expresses the group multiplication andǫthe group identity element e. Thus,if f∈C[G]the coalgebra is(∆f)(g,h)=f(gh),∀g,h∈Gǫf=f(e)(11)at least for suitable f(or with suitable topological completions).In other words it expresses the group product G×G→G by a map in the other direction in terms of coordinate algebras. From yet another point of view∆simply makes the dual H∗also into an algebra.So a Hopf algebra is basically an algebra such that H∗is also an algebra,in a compatible way,which makes the axioms‘self-dual’.For everyfinite-dimensional H there is a dual H∗.Similarly in the infinite-dimensional case.It said that in the Roman empire,‘all roads led to Rome’.It is remarkable that several different ideas for generalising groups all led to the same axioms.The axioms themselves werefirst introduced(actually in a super context)by H.Hopf in1947in his study of group cohomology but the subject only came into its own in the mid1980s with the arrival from mathematical physics of the large classes of examples(as above)that are neither like U(g)nor like C[G],i.e.going truly beyond Lie theory or algebraic group theory.Acknowledgements.An announcement of this article appears in a short millennium article[15] and a version more focused on the meaning for Planck scale physics in[16].II Quantum groups and Planck scale physicsThis section covers quantum groups of the bicrossproduct type coming out of Planck-scale physics[6]and their associated noncommutative geometry.These are certainly less well-developed than the more familiar U q(g)in terms of their concrete applications;one does not have inter-esting knot invariants etc.On the other hand,these quantum groups have a clearer physical meaning as models of Planck scale physics and are also technically easier to construct.Therefore they are a good place to start.Obviously if we want to unify quantum theory and geometry then a necessaryfirst step is to cast both in the same language,which for us will be that of algebra.We have already mentioned that vectorfields can be thought of classically as derivations of the algebra of functions on the manifold,and if one wants points they can be recovered as maximal ideals in the algebra, etc.This is the more of less standard idea of algebraic geometry dating from the late19th century and early on in the20th.It will certainly need to be modified before it works in the noncommutative case but it is a starting point.The algebraic structure on the quantum side will need more attention,however.II.A CogravityWe begin with some very general considerations.In fact there are fundamental reasons why one needs noncommutative geometry for any theory that pretends to be a fundamental one.Since gravity and quantum theory both work extremely well in their separate domains,this comment refers mainly to a theory that might hope to unify the two.As a matter of fact I believe that, through noncommutative geometry,this‘holy grail’of theoretical physics may now be in sight.Thefirst point is that we usually do not try to apply or extend our geometrical intuition to the quantum domain directly,since the mathematics for that has traditionally not been known. Thus,one usually considers quantisation as the result of a process applied to an underlying classical phase space,with all of the geometrical content there(as a Poisson manifold).But demanding any algebra such that its commutators to lowest order are some given Poisson bracket is clearly an illogical and ill-defined process.It not only does not have a unique answer but also it depends on the coordinates chosen to map over the quantum operators.Almost always one takes the Poisson bracket in a canonical form and the quantisation is the usual CCR or canonical commutation relations algebra.Maybe this is the local picture but what of the global geometry of the classical phase space?Clearly all of these problems are putting the cart before the horse:the real world is to our best knowledge quantum so that should comefirst.We should build models guided by the intrinsic(noncommutative)geometry at the level of noncommutative algebras and only at the end consider classical limits and classical geometry(and Poisson brackets)as emerging from a choice,where possible,of‘classical handles’in the quantum system.In more physical terms,classical observables should come out of quantum theory as some kind of limit and not really be the starting point;in quantum gravity,for example,classical geometry should appear as an idealisation of the expectation value of certain operators in certain states of the system.Likewise in string theory one starts with strings moving in classical spacetime, defines Lagrangians etc.and tries to quantise.Even in more algebraic approaches,such as axiomatic quantumfield theory,one still assumes an underlying classical spacetime and classical Poincar´e group etc.,on which the operatorfields live.Yet if the real world is quantum then phase space and hence probably spacetime itself should be‘fuzzy’and only approximately modeled by classical geometrical concepts.Why then should one take classical geometrical concepts inside the functional integral except other than as an effective theory or approximate model tailored to the desired classical geometry that we hope to come out.This can be useful but it cannot possibly be the fundamental‘theory of everything’if it is built in such an illogical manner.There is simply no evidence for the assumption of nice smooth manifolds other than now-discredited classical mechanics.And in certain domains such as,but not only,in Planck scale physics or quantum gravity,it will certainly be unjustified even as an approximation.Next let us observe that any quantum system which contains a nonAbelian global symmetry group is already crying out for noncommutative geometry.This is in addition to the more obvious position-momentum noncommutativity of quantisation.The point is that if our quantum system has a nonAbelian Lie algebra symmetry,which is usually the case when the classical system does, then from among the quantum observables we should be able to realise the generators of this Lie algebra.That is,the algebra of observables A should contain the algebra generated by the Lie algebra,A⊇U(g).(12)Typically,A might be the semidirect product of a smaller part with external symmetry g by the action of U(g)(which means that in the bigger algebra the action of g is implemented by the commutator).This may soundfine but if the algebra A is supposed to be the quantum analogue of the‘functions on phase space’,then for part of it we should regard U(g)‘up side down’not as an enveloping algebra but as a noncommutative space with g the noncommutative coordinates.In other words,if we want to elucidate the geometrical content of the quantum algebra of observables then part of that will be to understand in what sense U(g)is a coordinate algebra,U(g)=C[?].(13)Here?cannot be an ordinary space because its supposed coordinate algebra U(g)is noncom-mutative.。

Noncommutative Maxwell-Chern-Simons theory, duality and a new noncommutative Chern-Simons t

Noncommutative Maxwell-Chern-Simons theory, duality and a new noncommutative Chern-Simons t

a r X i v :h e p -t h /0302074v 2 24 M a r 2003hep-th/0302074Noncommutative Maxwell–Chern–Simons theory,duality and a new noncommutative Chern–Simons theory in d=3¨Omer F.DAYI 1The Abdus Salam ICTP,Strada Costiera 11,34014,Trieste–ItalyPhysics Department,Faculty of Science and Letters,Istanbul Technical University,80626Maslak–Istanbul,Turkey.Feza G¨u rsey Institute,P.O.Box 6,81220C ¸engelk¨o y–Istanbul,Turkey.AbstractNoncommutative Maxwell–Chern–Simons theory in 3–dimensions is defined in terms of star product and noncommutative fields.Seiberg–Witten map is employed to write it in terms of ordinary fields.A parent action is introduced and the dual action is derived.For spatial noncommutativity it is studied up to second order in the noncommutativity parameter θ.A new noncommutative Chern–Simons action is defined in terms of ordinary fields,inspired by the dual action.Moreover,a transformation between noncommuting and ordinary fields is proposed.1IntroductionAn equivalence of“ordinary”(commutative)and noncommutative gaugefields leads to a transformation between them which is known as Seiberg–Witten(SW)map[1].This permits one to study noncommutative gauge theories in terms of ordinaryfields.In fact, in[2](S)duality is incorporated into noncommutative Maxwell theory action in terms of ordinaryfields after performing SW map.In4–dimensions if the original theory is noncommutative Maxwell theory where noncommutativity is spatial,its dual theory is a noncommutative gauge theory whose time variable is effectively noncommuting with the other coordinates[2].This interesting phenomenon is a consequence of the fact that the duality transformation includes4–dimensional totally antisymmetric tensor.In3–dimensions the most extensively studied duality is between Maxwell–Chern–Simons(MCS)theory and self dual theory[3].It leads to two equivalent descriptions of the dynamics of massive spin–1field.One of its most known applications is bosoniza-tion in3–dimensions[4].We wonder what would be the consequences of generalization of this duality to noncommutative MCS theory.In[5]and[6]some generalizations of the mentioned duality to noncommutative theories are investigated in terms of noncommuting fields.However,duality can also be studied employing ordinaryfields in the spirit of[2]. Although atfirst sight this can appear to be trivial due to the fact that3–dimensional noncommutative Chern–Simons(CS)action becomes the usual CS action in terms of SW map[7]–[10],we will show that it gives nontrivial results.We write3–dimensional noncommutative MCS action in terms of ordinary gaugefields utilizing SW map.We introduce a parent action in terms of ordinaryfields to obtain the dual description.We study the dual action up to the second order in the noncommutativity parameterθ,when we let only spatial noncommutativity.Once the dual description is obtained it inspires a new noncommutative CS theory in terms of ordinary gaugefields. We discuss equations of motion following from this new action.Moreover,we propose to write it in terms of noncommutingfields as the simplest generalization of abelian CS action.This leads to an explicit transformation between noncommutative and ordinary fields.2Duality and Noncommutative MCS Theory It is well known that noncommutativity between coordinates can be introduced in terms of the star productiθµν∗≡expǫµνρ d3x ˆAµ∂νˆAρ+22and the noncommutative Maxwell theoryˆS M =−12Aρ∂µAν).(8) When the change of variables which follows from(7)is performed the noncommutative CS action(3)becomes the usual action[7]ˆS CS =m4[FµνFµν+L(θ,F)]+m8(θµνFµν)2FρσFσρ+12BµBµ+m4[Lθ(F)+Lθ2(F)+···] .(13)3Equations of motion with respect to B µareB µ=ǫµνρ∂νA ρ.When we substitute B µwith this in (13)the noncommutativeMCSaction(10)follows.On the other hand the equations of motion with respect to A µ∂ν ǫµνρ(B ρ−mA ρ)−1δF νµ=0,(14)can be solved for A µasA µ=1m b µ(θ,B ).(15)We defined b µ(θ,B )in terms of the equationb µ(θ,B )+1m +h (θ,B )δF µν.Obviously,b µ(θ,B )can be expanded in powers of θasb µ(θ,B )=b µθ+b µθ2+···.When we plug the solution (15)into (13)the dual of noncommutative MCS action follows:S D = d 3x 12mǫµνρ[B µ∂νB ρ+b µ(θ,B )∂νb ρ(θ,B )]−1m+h (θ,B )m 2[H µνH µν+2H 12H 12],b 1θ=2θm 2H 12H 10.4When we use these in(17)explicit form of the dual action to the second order inθfollows. To the second order inθ(17)can be written asS D,(2)≡ d3x 12mǫµνρBµ∂νBρ−1m +1Lθ2 H m2θµν(HνρhθρσHσµ+2HνρHρσhσµθ)412−2 d3xǫµνρBµ∂νBρ,(21) where M≡1/m.We would like to take advantage of this observation to define a new noncommutative abelian CS theory in terms of the ordinary gaugefields BµasS NCS= d3x M4L(θ,MH+Mh(θ,B)) ,(22)by dropping the B2term in(17).Obviously,this action is invariant under the abelian gauge transformationsδBµ=∂µλand yields the ordinary CS theory(21)when one sets θ=0.Equations of motion areǫµνρ∂ν(Bρ+bρ(θ,B)))−4ǫσνρ∂κ Gσκµ(H)∂νbρ =0,(23)where we definedδbµ(θ,B(y))Observe that the simplest solution of(23)isHµν=0,which is independent ofθ.To thefirst order inθthe equations of motion(23)get the simple form)=0.(25)ǫµνρ∂ν(Bρ+bρθThe SW map(7)expresses the noncommutative gaugefieldsˆAµin terms of the or-dinary gaugefields Aµutilizing the equivalence relation(8).A transformation between noncommutative and ordinaryfields can also be derived by assuming an equivalence rela-tion between the action(22)and another one written by introducing somefields B(B,θ) taking values in noncommutative space.However,there is no unique choice for the latter action.One should make an assumption about the form of the action in terms of the non-commutingfields B(B,θ).Let us suppose that the action in terms of the noncommutative fields B(B,θ),is in the same form as the abelian CS theory:MS NCS≡H21ǫµνρHνρ.2Although the assumption(26)is very plausible,in principle one may define some other actions in terms offields taking values in noncommutative space.Nevertheless,the assumed form of the action(26)is shown to yield a map between the noncommutative gaugefields Bµ(B,θ)and the ordinary ones Bµwhich is not the SW map(8).Moreover, the form of the action(26)can be useful to generalize this construction to nonabelian gauge theories.Acknowledgment:I would like to thank referee for useful comments.6References[1]N.Seiberg and E.Witten,JHEP9909(1999)032,hep-th/9908142.[2]O.J.Ganor,G.Rajesh and S.Sethi,Phys.Rev.D62(2000)125008,hep-th/0005046.[3]S.Deser and R.Jackiw,Phys.Lett.B139(1984)371.[4] E.Fradkin and F.A.Schaposnik,Phys.Lett.B338(1994)253,hep-th/9407182.[5]S.Ghosh,Gauge invariance and duality in the noncommutative plane,hep-th/0210107.[6]M.B.Cantcheffand P.Minces,Duality between noncommutative YMCS and non–abelian self–dual model,hep-th/0212031.[7]N.Grandi and G.A.Silva,Phys.Lett.B507(2001)345,hep-th/0010113.[8] A.P.Polychronakos,Annals of Phys.301(2002)174,hep-th/0206013.[9]K.Kaminsky,Y.Okawa and H.Ooguri,Quantum aspects of SW map in noncommu-tative CS theory,hep-th/0301133.[10]R.Banerjee,Anomalies in noncommutative gauge theories,SW transformations andRR couplings,hep-th/0301174.7。

纹理物体缺陷的视觉检测算法研究--优秀毕业论文

纹理物体缺陷的视觉检测算法研究--优秀毕业论文

摘 要
在竞争激烈的工业自动化生产过程中,机器视觉对产品质量的把关起着举足 轻重的作用,机器视觉在缺陷检测技术方面的应用也逐渐普遍起来。与常规的检 测技术相比,自动化的视觉检测系统更加经济、快捷、高效与 安全。纹理物体在 工业生产中广泛存在,像用于半导体装配和封装底板和发光二极管,现代 化电子 系统中的印制电路板,以及纺织行业中的布匹和织物等都可认为是含有纹理特征 的物体。本论文主要致力于纹理物体的缺陷检测技术研究,为纹理物体的自动化 检测提供高效而可靠的检测算法。 纹理是描述图像内容的重要特征,纹理分析也已经被成功的应用与纹理分割 和纹理分类当中。本研究提出了一种基于纹理分析技术和参考比较方式的缺陷检 测算法。这种算法能容忍物体变形引起的图像配准误差,对纹理的影响也具有鲁 棒性。本算法旨在为检测出的缺陷区域提供丰富而重要的物理意义,如缺陷区域 的大小、形状、亮度对比度及空间分布等。同时,在参考图像可行的情况下,本 算法可用于同质纹理物体和非同质纹理物体的检测,对非纹理物体 的检测也可取 得不错的效果。 在整个检测过程中,我们采用了可调控金字塔的纹理分析和重构技术。与传 统的小波纹理分析技术不同,我们在小波域中加入处理物体变形和纹理影响的容 忍度控制算法,来实现容忍物体变形和对纹理影响鲁棒的目的。最后可调控金字 塔的重构保证了缺陷区域物理意义恢复的准确性。实验阶段,我们检测了一系列 具有实际应用价值的图像。实验结果表明 本文提出的纹理物体缺陷检测算法具有 高效性和易于实现性。 关键字: 缺陷检测;纹理;物体变形;可调控金字塔;重构
Keywords: defect detection, texture, object distortion, steerable pyramid, reconstruction
II

Renormalisability of the matter determinants in noncommutative gauge theory in the envelopi

Renormalisability of the matter determinants in noncommutative gauge theory in the envelopi

a rXiv:076.452v1[he p-th]27J un27FTI/UCM 75-2007Renormalisability of the matter determinants in noncommutative gauge theory in the enveloping-algebra formalism C.P.Mart´ın 1and C.Tamarit 2Departamento de F´ısica Te´o rica I,Facultad de Ciencias F´ısicas Universidad Complutense de Madrid,28040Madrid,Spain We consider noncommutative gauge theory defined by means of Seiberg-Witten maps for an arbitrary semisimple gauge group.We compute the one-loop UV divergent matter contributions to the gauge field effective ac-tion to all orders in the noncommutative parameters θ.We do this for Dirac fermions and complex scalars carrying arbitrary representations of the gauge group.We use path-integral methods in the framework of dimen-sional regularisation and consider arbitrary invertible Seiberg-Witten maps that are linear in the matter fields.Surprisingly,it turns out that the UV divergent parts of the matter contributions are proportional to the noncom-mutative Yang-Mills action where traces are taken over the representation of the matter fields;this result supports the need to include such traces in the classical action of the gauge sector of the noncommutative theory.PACS:11.10.Nx;11.10.Gh;11.15.-qKeywords:Renormalisability,Seiberg-Witten map,noncommutative gauge theories.The issue of renormalisability of noncommutative gauge theories in the enveloping-algebraapproach has been a subject of intense research in the last years[1,2,3,4,5,6,7,8,9].Theoutcome of this research so far shows that NC Yang Mills is one-loop renormalisable up tofirst order inθ[5,9]—in fact,up to order two for the case of NC U(1)Yang-Mills[3]—butrenormalisability is spoiled by the presence of Dirac fermions in the fundamental repesentationor complex scalars in the U(1)case[2,3,4,8].However,in all cases the gauge sector of thetheory remains renormalisable despite the presence of matter;this has also been checked for anoncommutative extension of the Standard Model[6]in which the traces in the gauge sector aretaken over all the different particle representations.This renormalisability of the gauge sectoris quite intriguing and far from trivial since BRS invariance and power-counting do not accountfor it.Indeed,take a simple compact gauge group,then,power-counting and BRS invariancedo not restrict the one-loop UV divergent part of the effective action of the gaugefield in thebackground-field gauge to the noncommutative Yang-Mills action,but to a linear combinationwith arbitrary UV divergent coefficients of the noncommutative Yang-Mills action and termslikeθαβTr d4x Fαβ⋆Fµν⋆Fµν,θαβTr d4x Fµα⋆Fβν⋆Fµν,etc...The confirmation of the renormalisablity we have mentioned at higher orders inθand its understanding–perhaps,asa by-product of an as yet undiscovered symmetry of the theory–,as well as the study of itsdependence on the choice of traces for the noncommutative Yang-Mills action,are still openproblems.As afirst step in this direction,in this paper,we compute to all orders inθthe UV part ofthe one-loop effective action obtained by integrating out the matterfields in noncommutativegauge theory for arbitrary semisimple gauge group.By“matter”we mean Dirac fermions andcomplex scalars in an arbitrary unitary irreducible representation of the gauge group.Whatwe have obtained is that,in both cases and in dimensional regularisation with D=4+2ǫ,the pole part of the effective action of the gaugefield turns out to be proportional to thenoncommutative Yang-Mills action with traces taken over the representation of the gaugegroup acting on the matterfields,namely:=1Γf[A]one-looppole,A-dep.Tr d4x Fµν⋆Fµν,Fµν=∂µAν−∂νAµ−i[Aµ,Aν]⋆.(1)192π2ǫThis result supports the need to consider such types of traces for models aiming to have therenormalisability property;in fact,all models considered so far with a one-loop,order-θrenor-malisable gauge sector have these types of traces.A relevant example is the noncommutativeversion of the Standard Model in ref.[6],whose gauge sector involves a non-trivial sum oftraces over all the particle representations of the model.Our result holds for the class ofSeiberg-Witten applications for which the map between the noncommutative and ordinary matterfields is linear and invertible;the computation to all orders inθis feasible due to the possibility of changing variables in the functional integrals from the ordinaryfields to the noncommutativefields.The result we have obtained is quite surprising since BRS invariance and power-counting of the theory formulated in terms of the ordinaryfields do not enforce it,and,it is relevant in the phenomenological applications of noncommutative gauge theories, since it supports the robustness of the predictions based on the gauge sector of the theory. These phenomenological predictions can certainly be tested at the LHC[10,11,12,13].Using a similar notation as that employed in ref.[14],we consider a noncommutative gauge theory with a semisimple gauge group of the form G1×···×G N with G i simple for i=1...s and abelian for i=s+1,...,N.Then the ordinary gaugefield will be of the formaµ=sk=1g k(a kµ)a(T k)a+Nl=s+1g l a lµT l,where the T′s are generators of unitary irreducible representations of the group factors.The matterfields to consider are Dirac fermionsψand complex scalarsφin an irreducible rep-resentation of G and therefore carrying multi-indices I=i1...i s for the irreducible factors. In multi-index notation we can define generators and a“global”trace Tr as follows(T k)a IJ=δi1j1···(T k)a ik j k···δis j s,k=1,...,s,T l IJ=δi1j1···δis j sY l,l=s+1,...,N,Tr(T k)a(T k′)a′=(T k)a IJ(T k′)a′JI.In order to build noncommutative actions for the matterfields we need the Seiberg-Witten maps for the noncommutative gaugefield Aµand the matterfields,i.e.,noncommutative fermionsΨαI and complex scalarsΦI.We make no assumption on the map for the gauge field,but for the matterfields we consider maps of the formΨαI=(δIJδαβ+M[aµ,∂,γ;θ]αβIJ)ψβJ,ΦI=(δIJ+N[aµ,∂;θ]IJ)φJ,(2)and analogously for the Dirac adjoint fermion¯Ψ=¯Ψ†γ0and the complex conjugateΦ∗of Φ.γµdenotes the gamma matrices satisfying{γµ,γν}=2ηµν.With this notation we define next the actions for the matterfields in noncommutative spacetimeS f= d4x(¯Ψ⋆iγµDµΨ−m¯Ψ⋆Ψ),S sc= d4x((DµΦ)∗⋆DµΦ−m2Φ∗Φ)−V(Φ),(Dµ)IJ=δIJ∂µ−iAµIJ⋆,(3)where⋆denotes the usual Moyal product and V(Φ)is an arbitrary noncommutative gauge-invariant potential that will not contribute to the Aµ-dependent part of the gauge effective action.Our objective is to compute the divergent part of the one-loop effective actions for the gaugefield that are formally defined byΓf/sc[a;θ]=−i ln Z f/sc[a;θ],(4)Z f[a;θ]=N f [d¯ψ][dψ]exp(iS f),Z sc[a;θ]=N sc [dφ∗][dφ]exp(iS sc),N f/sc=Z f/sc[0;θ]−1.The previous expressions relate the gauge effective actions to the determinants of the operators appearing in the actions for the matterfields.We will make sense out of these formal definitions by using dimensional regularisation in D=4+2ǫdimensions;this will make the divergent contributions appear as poles inǫ.Note that,in order to work out the UV divergence ofΓf/sc to all orders inθby integrating out the matterfields in the functional integrals in eq.(4), one would need to know the Seiberg-Witten map to all these orders.We can avoid this by performing a change of variables in the functional integrals from the ordinaryfields to the noncommutative ones.Recalling eq.(2),[d¯ψ][dψ]=det(I I+M)det(I I+¯M)[d¯Ψ][dΨ],[dφ∗][dφ]=det(I I+N)−1det(I I+N∗)−1[dΦ∗][dΦ].The above determinants are defined in dimensional regularisation by a diagrammatic expansion where the propagators are equal to the identity.As a consequence they are given in momentum space by tadpole-like integrals,which are zero.Therefore,Z f[a;θ]=N f [d¯Ψ][dΨ]exp(iS f),Z sc[a;θ]=N sc [dΦ∗][dΦ]exp(iS sc)(5) and,since the matter actions given in eq.(3)depend on the noncommutative gaugefield Aµ, we have that the dependence of the effective actionsΓon aµis through a dependence on Aµ:Γf/sc[a;θ]=Γf/sc[A;θ].In particular,when the eqs.(4)and(5)definingΓf/sc are interpreted diagramatically,it is clear that the potential V(Φ)makes no contribution to the Aµ-dependent part of the gauge effective actions.Thus,eqs.(4)and(5)allow us to obtain the Aµ-dependent parts of the effective actionsΓf/sc as the determinants of the operators that appear in the actions for the matter in terms of the noncommutativefields.Integrating over[d¯Ψ],[dΨ]and[dΦ∗],[dΦ]neglecting V(Φ)we obtain:iΓf[A]A-dep.=ln det[∂/+im−iA/⋆]nTr[(∂/+im)−1iA/⋆]n,iΓsc[A]A-dep.=−ln det[iD2+im2]nTr[(i∂2+i m2)−1((∂·A)⋆+2A·∂⋆−iAµ⋆Aµ⋆))]n.(6)In the previous expressions Tr denotes a trace over discrete indices and integration over the continuous indices of the corresponding operators.The operators(∂/+im)−1and(i∂2+i m2)−1 have matrix elements given by ordinary propagators:y|(∂/+im)−1|x = d D p p2−m2+i0+, y|(i∂2+im2)−1|x = d D p p2−m2+i0+.(7) Since we are interested in the divergent part ofΓf/sc,we need to identify the contributions in eq.(6)that yield the poles at D=4.This can be done by using power-counting arguments as follows.Let usfirst note that the propagators in eq.(7)are diagonal in the colour indices, then,one realises that the trace in eq.(6)forces a trace in the colour indices of the background gaugefields AµIJ.Therefore we can writeiΓf/sc[A]A-dep.≡∞n=1d d x1... d d x n Tr[Aµ1(x1)...Aµn(x n)]Γf/sc(n)µ1...µn[x1,...,x n],(8)where the mass dimension ofΓf/sc(n)µ1...µn[x i]is,for D=4,4+3n.In momentum space Γf/sc(n)µ1...µn[p i]will be of dimension4−n and,due to the translation invariance of the propagators in eq.(7),it is given by a single loop integral.Then,power-counting tell us that Γf/sc(n)with n≥5arefinite.Thus we only need to work out contributions with up to4 background gaugefields Aµ.Let us start with the case of fermions.Going over to momentum space and starting from eq.(6),we can expressΓf(n)[A]in terms of a loop integral as followsΓf(n)µ1...µn [x1,...,x n]=(−1)n+1(2π)D(2π)Dδ i p i e i P n i=1p i x i e−i P n i<j p i◦p j×× dq[(q+p1)−m2][q2−m2][(q−p2)2−m2]···[(q− n−1i=2p i)2−m2].(9) p◦q≡ito the presence of interaction vertices with one and two A ′µs .Instead of a closed formula for Γ(n )as the one just given for fermions in eq.(9),we provide formulae for the potentially divergent contributions Γsc (k ),1≤k ≤4.Γsc (n )µ1...µn [x 1,...,x n ]= n i =1d D p i(2π)D2q µ(2π)D e −p 1◦p 2ηµν2 d D q [(q +p 1)2−m 2]2,˜Γ(3)µνρ[p 1,p 2,p 3]= d D q [q 2−m 2][(q −p 1)2−m 2]−1(2π)D e −P i<j p i ◦p j (2q +p 1)µ(2q −p 2)ν(2q −p 2+p 1)ρ2 d D q [q 2−m 2][(q −p 3−p 4)2−m 2]− d D q [q 2−m 2][(q −p 3)2−m 2][(q −p 3−p 4)2−m 2]+1(2π)De −P i<j p i ◦p j (2q +p 1)µ(2q −p 2)ν(2q −2p 2−p 3)ρ(2q −p 2−p 3+p 1)σReferences[1]A.Bichl,J.Grimstrup,H.Grosse,L.Popp,M.Schweda and R.Wulkenhaar,JHEP0106(2001)013[arXiv:hep-th/0104097].[2]R.Wulkenhaar,JHEP0203(2002)024[arXiv:hep-th/0112248].[3]M.Buric and V.Radovanovic,JHEP0210(2002)074[arXiv:hep-th/0208204].[4]M.Buric and V.Radovanovic,JHEP0402(2004)040[arXiv:hep-th/0401103].[5]M.Buric,tas and V.Radovanovic,JHEP0602(2006)046[arXiv:hep-th/0510133].[6]M.Buric,V.Radovanovic and J.Trampetic,JHEP0703,030(2007)[arXiv:hep-th/0609073].[7]X.Calmet,Eur.Phys.J.C50(2007)113[arXiv:hep-th/0604030].[8]C.P.Martin, D.Sanchez-Ruiz and C.Tamarit,JHEP02(2007)065[arXiv:hep-th/0612188].[9]tas,V.Radovanovic and J.Trampetic,arXiv:hep-th/0703018.[10]M.Buric,tas,V.Radovanovic and J.Trampetic,Phys.Rev.D75(2007)097701.[11]A.Alboteanu,T.Ohl and R.Ruckl,Phys.Rev.D74(2006)096004[arXiv:hep-ph/0608155].[12]M.Mohammadi Najafabadi,Phys.Rev.D74(2006)025021[arXiv:hep-ph/0606017].[13]W.Behr,N.G.Deshpande,G.Duplancic,P.Schupp,J.Trampetic and J.Wess,Eur.Phys.J.C29(2003)441[arXiv:hep-ph/0202121].[14]F.Brandt,C.P.Martin and F.R.Ruiz,JHEP0307(2003)068[arXiv:hep-th/0307292].[15]C.P.Martin and D.Sanchez-Ruiz,Nucl.Phys.B572(2000)387[arXiv:hep-th/9905076].。

Nonrepeatable GaugeR&R Studies Assuming Temporal or Patterned Object Variation

Nonrepeatable GaugeR&R Studies Assuming Temporal or Patterned Object Variation

Nonrepeatable Gauge R&R Studies Assuming Temporal or PatternedObject VariationFRANK VAN DER MEULENDelft University of Technology,Mekelweg4,2628CD Delft,The NetherlandsHENK DE KONINGA.T.Kearney,B.V.Van Heuven Goedhartlaan13,P.O.Box22926,1181LE Amstelveen,1100DK Amsterdam,The NetherlandsJEROEN DE MASTInstitute for Business and Industrial Statistics of the University of Amsterdam(IBIS UvA), Plantage Muidergracht12,1018TV Amsterdam,The NetherlandsThe standard method to assess a measurement system’s precision is a gauge repeatability and repro-ducibility(gauge R&R)study.It exploits replications to estimate variance components that are interpreted as measurement spread.For nonrepeatable measurements,it is not feasible to obtain replications because objects are destroyed when they are measured or because the object changes over time.Possible solutions are to replace replications with measurements of multiple objects or with the measurement of one object at multiple times.Subsequently,these measurements are modeled by afixed pattern(over time or over positions).We show that the experimental design used in this type of nonrepeatable gauge R&R studies is best constructed in a way that is similar to a Latin square design.These designs have a greatflexibility, can be applied in many situations encountered in practice,and have nice mathematical properties as well.We consider several examples in which this approach is applied and worked out.For the examples given, we provide the analysis and the results following the worked-out approach.Analysis of the envisaged exper-imental set-up is done with linear and nonlinear mixed models in which variance components are estimated by restricted maximum-likelihood estimators.Key Words:Latin-Square Experimental Design;Measurement Error;Nonlinear Mixed Model;Restricted Maximum Likelihood;Variance Components.T HE STANDARD method to assess a measurement system’s precision is a gauge repeatability andDr.Van der Meulen is an Assistent Professor at Delft Insti-tute of Applied Mathematics(Delft University of Technology). His e-mail adress is f.h.vandermeulen@tudelft.nl.Dr.De Koning is an associate consultant with A.T.Kear-ney.His e-mail address is Henk.deKoning@.Dr.de Mast is a principal consultant at IBIS UvA,and Associate Professor at the University of Amsterdam.He is a senior member of ASQ.His email address is j.demast@uva.nl.reproducibility(gauge R&R)study(see e.g.,Mont-gomery(2005),Burdick et al.(2003)).An example of the standard layout of such a study is presented in Table1.Each object out of a sample of objects is mea-sured multiple times by a number of operators.Vari-ation within rows is measurement spread.We denote the data by y ijk,where i indexes objects,j indexes operators,and k indexes replications.The data are modeled asy ijk=μ+a i+b j+(ab)ij+εijk.(1)2FRANK VAN DER MEULEN,HENK DE KONING,AND JEROEN DE MASTTABLE 1.Standard Layout of Gauge R&R StudyOperatorObjects 1231y 111y 112y 121y 122y 131y 1322y 211y 212y 221y 222y 231y 232 (10)y 10,1,1y 10,1,2y 10,2,1y 10,2,2y 10,3,1y 10,3,2Here μdenotes the overall average,a i ∼N (0,σ2a )are random object effects,b j ∼N (0,σ2b )are randomoperator effects,and (ab )ij ∼N (0,σ2ab )represent object–operator interaction.The εijk ∼N (0,σ2)are error terms.All a i ,b j ,(ab )ij ,and εijk are as-sumed stochastically independent.One is typically interested in the repeatability σ2,the reproducibil-ity σ2b +σ2ab ,and the total measurement spread σm =σ2+σ2b +σ2ab .The standard approach exploits replications toestimate measurement spread.For some measure-ments,it is not feasible to obtain replications,for example,because objects are destroyed when they are measured or because the object being measured changes over time.Such measurements used to be called destructive,but are nowadays often referred to as nonrepeatable.De Mast and Trip (2005)give a precise,mathematical definition of the problem of gauge R&R studies for nonrepeatable measurements.This comes down to the following:Nonrepeatable measurements are measurements for which either of the following two conditions does not hold:(I)Temporal stability ,by which we mean that thereal value of the object does not change in time.For example,suppose we measure the temper-ature of a piece of metal after heating and we wish to determine the error in this mea-surement.Because the metal may cool offvery rapidly,the temperature is not stable in time and the condition of temporal stability is vio-lated.(II)Robustness against measurement .This condi-tion is violated when the object is destroyed or changed significantly during measurement.An example is measuring the strength of bis-cuits,in which the biscuits break as a result of the measurement.Another example is mea-suring the rate of dissolution of a tablet.Afterthe measurement,the tablet is (partially)dis-solved,so it has changed significantly.Performing a gauge R&R study for nonrepeat-able measurements is a fundamental problem be-cause such measurements cannot be repeated under entirely equal conditions.This problem has been a persistent problem in quality engineering.Although there is no structural solution to it,there are a num-ber of approaches that work in some cases.De Mast and Trip (2005)give an overview of seven such ap-proaches.One of these approaches works with an experi-mental layout similar to the one in Table 1,but one in which the rows contain measurements on different objects instead of measurements on the same object.This experimental layout necessarily confounds mea-surement spread with object-to-object variation.The usual estimators for measurement spread now esti-mate σ2m +σ2ainstead of σm .If object-to-object variation (σ2a )within rows is not negligible,this ap-proach gives an overestimation of the measurement spread.Although this is commonly the case,the ap-proach is still useful because the bias is on the con-servative side:if the estimated measurement spread is acceptable,then the true measurement spread is as well.As suggested in De Mast and Trip (2005),this ap-proach can be improved upon if the object-to-object spread within rows is not just noise,but has a pat-tern,i.e.,if either of the following conditions hold:•Patterned temporal variation (PTV):the vari-ation over time of each object follows a certain pattern.•Patterned object variation (POV):the variation across objects follows a certain pattern.The idea is to fit a model for this systematic part of the within-rows-objects variation (condition POV)or temporal variation (condition PTV)and correct the data for it.This approach leads to a smaller over-estimation of measurement spread.Because we cor-rect for systematic differences between the objects within a row,the estimators for measurement spread (2m 2a )will be closer to σm .The approaches outlined above require a more ad-vanced experimental set-up and analysis than stan-dard gauge studies.This paper considers three cases,which are each introduced by a practical example:1.I is violated,but PTV holds.Objects vary overNONREPEATABLE GAUGE R&R STUDIES ASSUMING TEMPORAL OR PATTERNED OBJECT VARIATION3time,but the variation over time follows a cer-tain pattern that can be modeled.2.II is violated,but POV holds.Objects changeduring measurement,but their variation followsa pattern that can be modeled.3.In the third example,we consider the analy-sis of a dissolution testing gauge R&R experi-ment that is discussed in Gao(2007).Contraryto their analysis,we suggest taking patternedtemporal variation into account.In each case,we show the usefulness of Latin-square–related designs.The present work extends the work by De Mast and Trip(2005).First,it gives three practical case examples,showing for what kind of sit-uations their approach can be used.Furthermore,we pay much more attention to the experimental design, explaining the rationale for using certain designs,and extend the classes of experimental designs used,in-creasing theflexibility of this approach.Finally,the statistical analysis of the experiments is explained in more detail than in De Mast and Trip(2005).The remainder of this article is organized as fol-lows.The next section introduces the three case ex-amples,describes the experimental design,and pro-vides the actual data for thefirst two examples.Sub-sequently,we discuss appropriate statistical models for the data and their analysis.In thefinal section, conclusions are drawn.Experimental Design and Datafor Three CasesIn this section,three case examples will be intro-duced.For each example,we will discuss the experi-mental set-up and data.Measuring the Core Temperature of a Food Product:Example for PTV without IA food product is baked until its core reaches a temperature of about80◦C.The core temperature is measured by inserting a digital thermometer into the product.Because heat is not distributed perfectly homogeneously over the product and the operators insert the thermometer by feel(aiming for the core), it is likely that random measurement error is sub-stantial.To estimate random measurement error,we could do a standard gauge R&R study.Each food spec-imen could be measured multiple times,but be-cause the product cools down quite rapidly(about 1.0◦C per minute),these repeated measurements would confound measurement spread with variation in the product’s true core temperature(condition I—temporal stability—is violated).The Constructed Design:Latin-Square–Type Designs Let us examine how to best analyze the measure-ment error in this case.(Later we will see that the designs used in the other cases are very similar.)As-sume that we measure at n time instances.Assume, furthermore,that n=q×r,where q denotes the number of operators and r the number of times each operator measures a certain object.(This assumption is not needed,as we will point out later.)Ideally,each operator would measure each object at each time in-stant.However,at any given time instant,a particu-lar food specimen can only be measured by one op-erator.Nonetheless,we can create an experimental design in which each operator measures at each time instant,though not always the same object.This can easily be accomplished by a Latin-square design.To set the stage,consider the case for which we have three objects(n=3)and each operator measures once(r=1).If we denote the operators by A,B, and C,then an example of a design satisfying our requirements is the Latin-square design given byt1t2t3object1A B Cobject2B C Aobject3C A BIf we want to have each object measured twice by each operator(r=2),we can add measurements at3additional time instances t4,t5,and ing another Latin-square design for the measurements at these times,we obtain the following design:t1t2t3t4t5t6 object1A B C A C Bobject2B C A C B Aobject3C A B B A CAny permutation of the columns of this design will yield a design that suits our purposes.In a permuted design still,each operator measures each object twice (r=2),and still each operator measures at all time instants.This feature is what makes Latin-square de-signs tin-square designs can generally4FRANK VAN DER MEULEN,HENK DE KONING,AND JEROEN DE MAST tin-Square–Type Design(Entries Indicate Operator)TimeSpecimen060120180240300I A C B A B CII A A B C B C III C C A B A B IV C B A A C BV B A C B C A VI B B C C A A be obtained from standard Latin squares via per-mutation of rows,columns,and labels.(We refer to chapter31of Neter et al.(1985)for this.)Typically,one wants to measure more than three objects.To obtain a design for the more general case, we can adjust the above procedure,starting with Latin squares of a different(higher)u-ally,we can set up the experiment in such a way that the condition n=q×r holds,but if this con-dition does not hold,an appropriate design can be constructed by deleting rows or columns in a larger design(see also Cochran and Cox(1957)and chap-ter31in Neter et al.(1985)about these so called “Youden”designs).Experimental Set-Up and DataIn the actual experiment,it was decided to select six specimens of the food product(n=6).Each specimen was to be measured twice(r=2)by each of three operators(q=3),according to the design in Table2.The two times three measurements were to be done with60seconds between successive measure-TABLE3.Measurements for Food-Product ExperimentTimeSpecimen060120180240300 I87.083.982.282.077.076.4 II78.176.875.073.869.270.4 III77.277.777.076.476.373.1 IV74.372.873.970.970.769.6 V81.681.979.979.378.378.1 VI77.676.174.873.974.274.2FIGURE1.Data,Food Experiment.ments.We constructed the Latin-square–type exper-imental design that is shown in Table2.The results of the experiment are shown in Table3.The entries are the measured core temperatures(◦C).Figure1 shows a Trellis graph of the core temperature over time per food product specimen.Before proceeding to the analysis of this exper-iment,wefirst introduce two other examples.The statistical analysis of the three examples is given in the upcoming section.Measuring Shrinkage of Carpet Tiles: Example for POV Without IIA company produces carpet tiles.Out of a stretch of carpet,carpet tiles are blanked.After production, the amount of shrinkage(or expansion)of the carpet tiles is measured.If the carpet tiles shrink or expand during their lifetimes,customers complain.The challenge in this measurement procedure is to mimic the circumstances to which the carpet tiles are exposed during their lifetime.In order to stress test the carpet tile’s performance with respect to shrinkage,it is exposed to extreme temperatures and moisture conditions during the measurement.The gauge R&R of such stress tests refer to the consis-tency of the test results that would be obtained if the test were performed multiple times on the same tile. However,these tests are irreversible and therefore the measurement procedure is nonrepeatable(von-dition II—robustness against measurement—is vio-lated here).NONREPEATABLE GAUGE R&R STUDIES ASSUMING TEMPORAL OR PATTERNED OBJECT VARIATION5 Experimental Set-Up and DataIn order to determine how to construct a good de-sign,we need to know a little bit more about the process of producing carpet tiles.As we explained, carpet tiles are blanked out of a stretch of carpet. From the left to right side of the the carpet stretch, six tiles are blanked.It is assumed,however,that the amount of shrinkage varies from the left to right side of the carpet stretch.So the shrinkage follows a pat-tern that can be modeled,similar to the pattern over time in thefirst case example(in which the temper-ature of a food product was measured).Therefore, we choose a design that is similar to the one used in the food experiment(see Table2).Note that,in this case,“time”should be interpreted as the position at which the carpet tile is blanked out of the stretch. Position1indicates the position at the utmost left, position6the position at the utmost right of the stretch of carpet.The results of the experiment are shown in Table 4.Entries are measured shrinkage percentages.Fig-ure2shows the shrinkage per position(from left to right)per carpet tile.Measuring Dissolution of Tablets:Example for PTV Without I and IIThe third case is based on an example given in a paper of Gao et al.(2007).They examine the process of dissolution testing of tablets.The rate at which tablets dissolve is an important aspect in pharma-ceutical applications.The dissolution rate of tablets is measured by an apparatus that exposes the tablets to theflow of some liquid.Over time,measurements are done.The drug released into the medium from the tablet matrix is measured.The measurement is, of course,nonrepeatable because the tablet is par-tially gone after the measurement.As we will see, TABLE4.Measurements for Carpet-Tile ExperimentPositionCarpet tile123456 I0.690.420.280.380.150.4II0.810.550.230.270.230.43 III0.670.450.420.210.360.38 IV0.530.280.320.220.120.23V0.510.430.200.190.240.41 VI0.470.290.220.230.270.45FIGURE2.Data,Carpet-Tile Experiment.the dissolution percentage of a single tablet follows a certain pattern in time.Experimental Set-Up and DataThe experiment of Gao(2007)is as follows. Two operators measure the dissolution percentage of tablets on each of two apparatuses(labeled A and B).Each apparatus contains six separate ves-sels.This experiment is then a three-factor crossed then nested model design.The whole experiment is replicated6times,resulting in6runs.A single mea-surement on one tablet consists of measurements at7 time instants(t=7.5,15,30,45,60,75,90minutes). Figure3shows these measurements for two typical tablets.Gao(2007)based the gauge R&R analysis onthe FIGURE3.Data,Tablets Experiment(Dissolution Profile for Two Tablets).6FRANK VAN DER MEULEN,HENK DE KONING,AND JEROEN DE MASTmeasurements conducted at time t=30andfit a model similar to the one specified in Equation(1). Our approach in this case is twofold.First,we will model the dissolution profile in time,thereby using all data available from the experiment.Gauge R&R results are obtained fromfitting a nonlinear mixed-effects model.This results in estimates of measure-ment error at all measurement times of the experi-ment.Second,we will suggest another experimental design(with fewer runs)that could have been used, resulting in reduced confounding of tablet and mea-surement error.Analyzing the Three Examples General Remarks:Estimation Method The standard gauge-R&R model,as given in Equation(1),is an example of a linear mixed model (mixed referring to the presence of both random and fixed effects).In analyzing thefirst two examples,we will use linear mixed models as well.For the third example,we willfit the data by a nonlinear mixed model.In this section,we make some comments on the way these models can be estimated and our pre-ferred method.The traditional way to estimate variance com-ponents is the ANOVA method.Unfortunately, ANOVA has some well-established drawbacks,espe-cially in the case of unbalanced data(which is im-portant for our examples).The most important are (see Searle et al.(1992),pp.35–39)1.The possibility of negative estimates for vari-ance components,which are not realistic froma practical viewpoint.2.The lack of uniqueness of the choice of sumsof squares.For unbalanced data,one can use,for instance,Henderson’s methods I,II,andIII,which are all using different sets of sumsof squares.On top of this,criteria for decidingwhich choice for sums of squares is optimal arelacking.Therefore,the choice for a particularset of sums of squares is arbitrary. Maximum likelihood(ML)estimation is a viable al-ternative for estimating variance components.Neg-ative estimates are impossible when using ML es-timation and the problem of the arbitrary nature of ANOVA is solved as well.An additional bene-fit of ML estimation is that the resulting estima-tors are asymptotically efficient.Detailed explana-tion of its application to variance–component esti-mation can be found in Searle et al.(1992).A slight drawback of ML estimation for mixed models(which includes the models under study)is that estima-tors for variance components depend on thefixed effects.This problem is circumvented by restricted maximum-likelihood(REML)estimation,where the fixed parameters are treated as nuisance parameters (see McCullagh and Nelder(1989),chapter8,or Da-vidian and Giltiman(1995),chapters3and4).The restricted likelihood is defined as the likelihood ob-tained by integrating out thefixed effects.By maxi-mizing the restricted likelihood over the set of vari-ance components,we obtain the REML estimators. Once the REML estimates are computed,they re-place the variance components in the(generalized) least squares equations.From these,we can then ob-tain estimates for thefixed effects.REML estimators can be justified from a Bayesian point of view as well(Searle et al.(1992),section 9.2b).In the Bayesian setup,besides the random ef-fects,also thefixed effects are considered random. If we use a noninformative improper(flat)prior dis-tribution on the elements offixed effects,then it is not hard to see that the resulting posterior density is proportional to the restricted likelihood(the propor-tionality constant not depending on the random ef-fects).As a consequence,the REML estimator equals the posterior mode if we use a noninformative prior for thefixed effects.In the following,we will use the“nlme”library of S-plus to perform the statistical analysis of the ex-amples.We will report REML estimates.For a de-tailed explanation of the nlme package,we refer to Pinheiro and Bates(2000).Analysis of ExamplesMeasuring the Core Temperature of a Food Product First wefit the data to the standard gauge R&R model of Equation(1).Although it is obvious from Figure1that the measured temperatures for each specimen are time dependent,it is instructive to com-pare this“naive”approach(neglecting time as co-variate)to a more sophisticated one,which we will outline below.Hence,denote by y ijk the temperature for specimen i,measured by operator j for the k th time.Fitting a crossed-random effect models yields the results summarized in Table5(Here and in the following,we report95%confidence intervals.In the output these are denoted by[lower,upper];est.de-notes the REML estimate.Furthermore,fixed effects are aligned to the left,random effects are indented.)NONREPEATABLE GAUGE R&R STUDIES ASSUMING TEMPORAL OR PATTERNED OBJECT VARIATION7 TABLE5.Results for Fitting a Crossed Random-Effects ModelLower Est.UpperIntercept73.3676.4379.47 Specimen 1.73 3.42 6.74 Operator0.050.617.23 Specimen:operator0.000.434521.91 Residual 1.81 2.49 3.44 The estimates for random effects are standard devia-tions.The width of the confidence interval for the in-teraction term is huge.Blind application of the stan-dard model is therefore not sensible,as opposed to what is often done in practice.An ANOVA test shows that the interaction term can be dropped from the model.Refitting gives the results of Table6.There-fore,the“naive”approach shows that reproducibility is a minor issue,but repeatability is a serious issue for this measurement.In fact,the estimated measure-ment spread equals2.60.The proportion of measure-ment variance due to repeatability equals94%.Next,we will start with a very easy model and refine it in steps.Looking at Figure1,it seems that, for each object,the temperature decreases linearly in time.Let t1,...,t6denote the times of measurement in the experiment.If y ik denotes the temperature of specimen i at time t k,wefit the modely ik=μ+γ(t k−150)+εik.(2) Here{εik}denotes a sequence of independent identi-cally distributed normal random variables.The dot plots of the residuals per specimen in Figure4re-veal that this model is too simplistic.Because the specimens are drawn from a population of specimens, it is natural to model their effects as random ef-fects.This is confirmed by Figure5,which showsTABLE6.Results for Fitting a Random-Effects Model Without InteractionLower Est.Upper Intercept73.3876.4379.47 Specimen 1.74 3.42 6.73 Operator0.060.62 6.55 Residual 1.93 2.523.27FIGURE4.Dot Plots of Residuals for Each Specimen for Naive Model.thefitted parameters plus confidence intervals in case we apply a linearfit for each specimen separately. The confidence intervals for the intercept parameter β0show little overlap.Additional drawbacks of the fixed-effects model are,first,that it does not pro-vide an estimate of the between-specimen variability (which we are interested in)and,second,that the number of parameters in the model grows linearly with the number of specimen.Therefore,we propose the following mixed analysis of covariance model: y ik=μ+a i+γ(t k−150)+εik.(3) Hereμandγarefixed effects and a i is a random specimen effect.To see if an operator effect should be taken into account as well,we plot the residu-als obtained byfitting model(3)in Figure6.We made separate box plots for each operator.Thisfig-ure points out a distinction between operator Aand FIGURE5.Confidence Intervals for Estimates in Simple Linear Fit for Each Specimen.8FRANK VAN DER MEULEN,HENK DE KONING,AND JEROEN DEMASTFIGURE 6.Residuals per Operator for Model (3).operators B and C.We will assume that the three operators are drawn from a population and therefore model its effect as a random effect.This leads to the modely ik =μ+αi +β +γ(t k −150)+εik ,(4)where indexes operators and {εik is a sequence of independent and identically distributed normal ran-dom variables.With this model,we obtain an esti-mate for the between-operator spread as well (adding an additional random interaction term gives similar problems as in the naive approach).The results for model (4)are summarized in Table 7.Further resid-ual plots did not reveal interesting patterns.We con-clude that the measurement spread σm equals 1.42and that 60%of the measurement variance is due to repeatability.This analysis clearly shows that the es-timated measurement spread is in fact much smaller than the naive analysis pointed out.Measuring Shrinkage of CarpetsThe analysis for this example is similar to that of the food product data.We fitted both the standardTABLE 7.Results for Fitting Model (4)LowerEst.Upper Intercept73.3876.4379.58I(time −150)−0.023−0.030−0.016Specimen 1.89 3.55 6.66Operator 0.300.90 2.70Residual0.841.101.44TABLE 8.Results for Fitting a Standard GaugeR&R Model Without Interaction TermLowerEst.Upper Intercept 0.280.400.52Batch 0.00110.01500.2134Operator 0.01630.06560.2640Residual0.11480.14670.1875gauge R&R model (without interaction term)and the following model:y ijk =μ+αi +βj +γk +εijk .(5)Here α,β,and εare independent batch,operator,and repeatability random effects,respectively;μis the overall mean and γ1,...,γ6are fixed position ef-fects.For identifiability,we will set γ1=0,which cor-responds to treatment contrasts.The results for the naive analysis are in Table 8.Therefore,σm ≈0.16.The proportion of measurement variance due to re-peatability equals 83%.The results for fitting model (5)are in Table 9.From this table,it follows thatσm ≈0.07662+0.05192≈0.093.The proportion of measurement variance due to re-peatability equals 46%.Again,we see that the over-estimation of the measurement spread is relatively large.Measuring Dissolution of TabletsWe now discuss the analysis of the gauge R&R experiment of Gao (2007).We are grateful to the au-TABLE 9.Results for Fitting Model (5)LowerEst.Upper Intercept 0.580.680.79Position2−0.27−0.21−0.15Position3−0.40−0.34−0.27Position4−0.42−0.36−0.30Position5−0.45−0.39−0.33Position6−0.29−0.23−0.17Batch 0.00460.02020.0880Operator 0.02770.07660.2120Residual0.03980.05190.0678NONREPEATABLE GAUGE R&R STUDIES ASSUMING TEMPORAL OR PATTERNED OBJECT VARIATION9thors of Gao(2007)for sharing their data with us.In dissolution testing,variability arises primarily from three factors:apparatus,operator,and tablet formu-lation and/or manufacturing process(Gao(2007),p. 1796).Therefore,the following model could befit for the30-minute data:y ijk=μ+b j+c k+(bc)jk+εijk.(6) Here y ijk denotes the dissolution percentage after30 minutes of tablet i,measured by operator j at ap-paratus k,b j denotes a random operator effect,c k a random apparatus effect,and(bc)jk a random opera-tor×apparatus interaction effect.However,because there are only two operators(apparatuses)involved in this experiment,we doubt whether it is sensible to treat them as randomly chosen from a possibly larger population of operators(apparatuses).Instead,we suggest the following model to include their effects: y ijk=μ+εijk,withεijk i.i.d.∼N(0,σ2jk),j=A,B,k=1,2.(7) Here j and k are index operator and apparatus,re-spectively.Model(7)is a short way to write down the model for which the measurements by operator A on apparatus1are assumed to follow a N(μ,σ2A1) distribution,measurements by operator A on appa-ratus2have a N(μ,σ2A2)distribution,measurements by operator B on apparatus1have a N(μ,σ2B1)dis-tribution,and measurements by operator B on ap-paratus2have a N(μ,σ2B2)distribution.As pointed out on page209in Pinheiro and Bates(2000),S-plus uses a different parametrization,in which it estimates σ2A1byˆσ2A1and expresses the estimates for the other variance parameters as multiples ofˆσ2A1.Tablet63 contains an exceptionally low value at time t=15 and is excluded from this analysis.The results for fitting model(7)are summarized in Table10.TABLE10.Each Line Shows the Results fromFitting Model(7)to the Dissolution Dataat a Certain Fixed TimeTime103σA1103σA2103σB1103σB27.513.9012.9217.60 5.40 1513.3912.1716.45 3.56 309.6810.9712.17 3.28 45 5.658.769.10 2.52 60 3.338.937.30 2.86 75 3.009.157.57 3.61 90 3.2110.017.97 4.41We will now deduce one statistical model for the data at all time instances.We start with a very simple model that specifies the dissolution profile in time.As we go along,we will include factors(sim-ilarly as in the analysis of the food-product experi-ment).Let y i denote the measurement at time t of the i th tablet.We propose to model the shape of the profile by the Weibull model,y i =f(t ;˜a,˜r,h)+εi ,f(t;˜a,˜r,h)=˜a1−exp−˜r t1−h1−h,t≥0.(8)Here{εi }is a sequence of independent and iden-tically distributed N(0,σ2ε)random variables.The parameters˜a and˜r are assumed to be positive and h<1.Note that this modeling step is analogous to the model in Equation(2)for the food-product ex-ample.The only difference is that here we do not pro-pose a linear function of time,but a nonlinear func-tion.Depending on the values of˜a,˜r,and h,we will see that the graph of the mapping t→f(t;˜a,˜r,h) looks like the curves observed in Figure3.Because h<1,lim t→∞f(t;˜a,˜r,h)=˜a,the parameter˜a models thefinal fraction of dissolved tablet.An interpretation for the other two parame-ters(˜r and h)is given by L´a nsky and Weiss(2003). We now briefly explain their approach.If we define C(t)=f(t;˜A,˜r,h)/˜a,then C(t)is the amount of dissolved tablet at time t,divided by the amount of tablet that is going to be dissolved ultimately,as time increases.L´a nsky and Weiss(2003)interpret C(t)as the cumulative distribution function of a ran-dom variable T,which is defined as the dissolution time of a randomly selected molecule(or the time until a molecule of the tablet enters solution).Thus, C(t)=P(T≤t).With this probabilistic interpreta-tion,we can give a meaning to the parameters˜r and h.Suppose T has a probability density f.Define the fractional dissolution rate byk(t):=f(t)1−F(t),t≥0.(9)From this definition,it is seen that the fractional dissolution rate is analogous to the hazard rate in survival analysis.Its probabilistic interpretation is as follows(see L´a nsky and Weiss(2003),p.1633): k(t)dt equals the probability that a molecule in solid state will be dissolved in the interval[t,t+dt)under the condition that this has not happened up to time t.For the Weibull model,the fractional dissolution。

  1. 1、下载文档前请自行甄别文档内容的完整性,平台不提供额外的编辑、内容补充、找答案等附加服务。
  2. 2、"仅部分预览"的文档,不可在线预览部分如存在完整性等问题,可反馈申请退款(可完整预览的文档不适用该条件!)。
  3. 3、如文档侵犯您的权益,请联系客服反馈,我们会尽快为您处理(人工客服工作时间:9:00-18:30)。
February, 2000
Abstract
We construct noncommutative gauge theories based on the notion of the Weyl bundle, which appears in Fedosov’s construction of deformation quantization on an arbitrary symplectic manifold. These correspond to D-brane worldvolume theories in non-constant B-field and curved backgrounds in string theory. All such theories are embedded into a “universal” gauge theory of the Weyl bundle. This shows that the combination of a background field and a noncommutative field strength has universal meaning as a field strength of the Weyl bundle. We also show that the gauge equivalence relation is a part of such a “universal” gauge symmetry.
∗ asakawa@gauge.scphys.kyoto-u.ac.jp † ikishimo@gauge.scphys.kyoto-u.ac.jp
1
Introduction
D-brane worldvolume theory in a constant B-field background is described by so-called noncommutative Yang-Mills theory, whose multiplicative product is the Moyal-Weyl product. This has been observed in the context of Matrix Theory in [3][4] and recently various research in this theory is discussed in string theory viewpoint.1 However, the more general situation, i.e., in a non-constant B-field background has not been understood yet. In this paper we propose one construction of such theories from the point of view of purely worldvolume theory. The idea is as follows: the Moyal-Weyl product appears originally in the deformation quantization of the Euclidean space R2n . This scheme is generalized to the quantization of any symplectic or Poisson manifold and resulting product is called the star product. If we regard this not as quantized space but as noncommutative geometry, a field theory with such product is the generalization of the noncommutative Yang-Mills theory. Although this idea appears extensively in the literature [7], explicit construction seems not to be made. The deformation quantization of a Poisson manifold (M, {, }) was first defined and investigated in [8]. Let Z = C ∞ (M )[[ ]] be a linear space of formal power series of the deformation parameter with coefficients in C ∞ (M ):
KUNS-1643 hep-th/0002138
Noncommutative Gauge Theories
arXiv:hep-th/0002138v6 22 Nov 2000
from Deformation Quantization
Tsuguhiko Asakawa∗and Isao Kishimoto† Department of Physics, Kyoto University, Kyoto 606-8502, Japan

f=
k=0
k
fk
(1)
Hale Waihona Puke Deformation quantization is an associative algebra structure on Z with some associative product ∗

f ∗g =
k=0
k
Mk (f, g )
(2)
where Mk are bidifferential operator such that M0 (f, g ) = f g, M1 (f, g ) − M1 (g, f ) = −i{f, g }. Two star products ∗1 and ∗2 are called equivalent if there exists an isomorphism of algebras T : (Z, ∗1 ) → (Z, ∗2 ) given by a formal power of differential operators T = T0 + T1 + · · · . Although we regard as some scale of noncommutativity, we use the term “quantum” as well. The problem of existence and classification up to above equivalences on an arbitrary symplectic manifold was solved by several authors [10][11]. Finally Kontsevich [8] solved in the case of an arbitrary Poisson manifold and relation with string theory was also suggested [9]. Above mentioned papers almost follow this line. We here prefer to consider the symplectic case only, because in this case, Fedosov [1][2] has given nice simple geometrical construction based on the Weyl algebras bundle. Let us recall his original method [1] briefly. Since each tangent space of a symplectic manifold is a symplectic vector space, it can be quantized by usual Moyal-Weyl product. These fibers constitute a bundle of algebras, which is a sort of “quantum tangent bundle.” Then Fedosov constructed a flat connection on it, adding some quantum correction to the usual affine connection. The flat sections of this connection can be naturally identified with Z = C ∞ (M )[[ ]]. So the product on fibers induces a star product on Z . In §2 we review some generalization of this method in detail.
2
Fedosov’s ∗ Product
In this section, we will recall Fedosov’s construction of ∗ products on an arbitrary symplectic manifold [1][2]. Here we deal with more general version [2], which will be necessary for our purpose.
1
There are many papers on this subject. For example, [5][6] and references therein.
1
In §3 we show that the automorphism of the Weyl algebras bundle is regarded as some (infinite dimensional) gauge transformation and this relates equivalent star products. Moreover, its subgroup, which preserve a star product, corresponds to the so-called noncommutative gauge transformation. From this observation we show that the Weyl algebra bundle itself has physical meaning as some infinite dimensional gauge bundle and noncommutative gauge field is naturally introduced by its restriction (in §4). We will construct explicitly such a gauge theory. As a result, the field strength of universal gauge field is a combination such as one in Born-Infeld action. We also show that two such gauge fields satisfy the gauge equivalence relation. Further physical implications of our noncommutative gauge theories are discussed in §5.
相关文档
最新文档