chunk_uid
stringlengths
40
40
chunk_type
stringclasses
2 values
chunk_index
int64
0
6.71k
total_chunks
int64
1
6.71k
section_title
stringlengths
1
157
embed_text
stringlengths
1
83.3k
spans
dict
paper_doi
stringlengths
0
63
paper_id_arxiv
stringlengths
9
16
title
stringlengths
7
245
authors
listlengths
1
768
categories
listlengths
1
7
year
int64
2k
2.02k
language
stringclasses
2 values
discipline
stringclasses
8 values
dense_vector
listlengths
1.02k
1.02k
e11ec2fc48e05cf5894e90f757b69771307e299f
subsection
80
84
Line-search: The Newton-secant solver
The details are left to the readerNewton-Secant solver [1] k = 0, k_{\max } \ge 1, \ell _{l} < \ell _{r}, g(\ell _{l}) > 0, g(\ell _{r}) < 0, \textup {tol} > 0 k \le k_{\max } and \ell _{r} - \ell _{l} > \textup {tol} k := k + 1 \ell _l^{\textup {aux}} := \ell _{l} g(\ell _{l}) > g(\ell _{r}) s_l := \frac{g(\ell _{r...
{ "cite_spans": [] }
10.1016/j.cma.2018.11.036
1807.02563
Invariant domain preserving discretization-independent schemes and convex limiting for hyperbolic systems
[ "Jean-Luc Guermond", "Bojan Popov", "Ignacio Tomas" ]
[ "math.NA" ]
2,018
en
Mathematics
[ -0.04438937082886696, -0.003353990614414215, -0.029836978763341904, -0.024986181408166885, 0.035267431288957596, -0.017206601798534393, 0.03664029762148857, 0.03630470857024193, -0.004797407891601324, 0.029775962233543396, -0.017221854999661446, 0.0099990488961339, -0.02573363110423088, 0....
601b7447a6a706324e8c94a37af355e0e01d53dc
subsection
81
84
Line-search: The Newton-secant solver
For instance, if the initial guess \ell ^0 \in [0,1] for Newton's method is such that \ell ^0 > \ell ^* (g(\ell ^0) < 0), then Newton's method produces a sequence \lbrace \ell ^k\rbrace _{k \in } satisfying \ell ^* < \ell ^k for all k\in . This implies that g(\ell ^k) < 0 for all k\in , which is incompatible with the c...
{ "cite_spans": [] }
10.1016/j.cma.2018.11.036
1807.02563
Invariant domain preserving discretization-independent schemes and convex limiting for hyperbolic systems
[ "Jean-Luc Guermond", "Bojan Popov", "Ignacio Tomas" ]
[ "math.NA" ]
2,018
en
Mathematics
[ -0.05571071803569794, 0.008298639208078384, -0.006033293902873993, -0.014743814244866371, 0.012348802760243416, 0.0021032628137618303, -0.009313086979091167, 0.000015016495126474183, -0.004332378040999174, 0.03999822214245796, -0.02157036028802395, 0.010686024092137814, -0.030936839058995247...
f20902cd841eb2c99c04df658c644487dd579492
subsection
82
84
Relaxing the bounds
In general the quantity \Psi _i^{\min } defined in (REF ) is accurate enough to make the limited high-order solution second-order in the L^1-norm in space. But it is too tight to make the method higher-order or even second-order in the L^\infty -norm in the presence of smooth extrema. The situation is even worse when u...
{ "cite_spans": [] }
10.1016/j.cma.2018.11.036
1807.02563
Invariant domain preserving discretization-independent schemes and convex limiting for hyperbolic systems
[ "Jean-Luc Guermond", "Bojan Popov", "Ignacio Tomas" ]
[ "math.NA" ]
2,018
en
Mathematics
[ -0.06335566937923431, 0.009987062774598598, -0.04052809625864029, -0.003936840686947107, -0.0070535060949623585, -0.0050354935228824615, 0.015602401457726955, -0.019150136038661003, 0.03111324831843376, 0.04348835349082947, -0.03820871561765671, 0.017349565401673317, -0.022766536101698875, ...
98c541174a800f66e4d57720b016ea7d83db12ee
subsection
83
84
Relaxing the bounds
The exponent 1.5 is somewhat ad hoc; in principle one could take r_i = (\frac{m_i}{||})^{\frac{\delta }{d}} with \delta <2.
{ "cite_spans": [] }
10.1016/j.cma.2018.11.036
1807.02563
Invariant domain preserving discretization-independent schemes and convex limiting for hyperbolic systems
[ "Jean-Luc Guermond", "Bojan Popov", "Ignacio Tomas" ]
[ "math.NA" ]
2,018
en
Mathematics
[ -0.04135126993060112, 0.0023746462538838387, -0.01707456447184086, 0.03344722464680672, -0.01988217793405056, -0.007438650820404291, -0.002738949377089739, 0.06088249757885933, 0.04708856716752052, -0.0006594649748876691, -0.04428095370531082, -0.00854491163045168, -0.027816738933324814, 0...
87190a34c5fd82003c63dabf35b5c650b285a535
abstract
0
25
Abstract
In this work, we address some optimal control problems related to the evolution of two isothermal, incompressible, immisible fluids in a two dimensional bounded domain. A distributed optimal control problem is formulated as the minimization of a suitable cost functional subject to the controlled nonlocal Cahn-Hilliard-...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ 0.0014987668255344033, -0.008283261209726334, -0.010891801677644253, 0.02707695960998535, -0.015895318239927292, 0.00018686914700083435, -0.020776798948645592, -0.018290294334292412, 0.0244073998183012, 0.06254395842552185, -0.02963973581790924, -0.03432290628552437, -0.005941676441580057, ...
f0ec1e55fd749b28a3945be271f1cff2f9984bdf
subsection
1
25
Body
lemmatheorem lemmapropositiontheorem propositionassumptiontheorem assumptioncorollarytheorem corollarydefinitiontheorem definitionexampletheorem exampleremarktheorem remarkhypothesistheorem hypothesispropertytheorem property[Pontryagin Maximun Principle and Second order optimality conditions]Pontryagin Maximun...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.022460253909230232, -0.00843022484332323, -0.024382803589105606, 0.020339347422122955, -0.009803473949432373, -0.00044177621020935476, -0.028487293049693108, 0.02082761377096176, -0.00005900680480408482, 0.02627483569085598, -0.044096559286117554, 0.010787636041641235, -0.0227043870836496...
71d6fa48c58d1b5c1171a41e9b640a6b52736287
subsection
2
25
Body
We describe the first order necessary conditions of optimality via Pontryagin minimum principle and prove second order necessary and sufficient conditions of optimality for the problem.Key words: optimal control, nonlocal Cahn-Hilliard-Navier-Stokes systems, Pontryagin maximum principle, necessary and sufficient optima...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.01175758522003889, 0.0009089048253372312, -0.025056397542357445, 0.001581284566782415, 0.019593432545661926, -0.007538283243775368, -0.020570050925016403, -0.005111993756145239, 0.027879439294338226, 0.05160824581980705, -0.027513207867741585, -0.01782330870628357, -0.010124798864126205, ...
2cb07a6fb14679cd0ba669749af2d5be9d7d6409
subsection
3
25
Body
In , author studies the local Cahn-Hilliard-Navier-Stokes system and establishes the existence of weak solutions in dimensions 2 and 3. The existence and uniqueness of strong solutions in 2 and 3 dimensions (global in 2D and local in time for 3D) is also established in . The authors in proved the existence of a weak so...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.019896341487765312, -0.008483408950269222, -0.014487405307590961, 0.014182246290147305, 0.00817062146961689, 0.026792926713824272, 0.02631993032991886, 0.021635744720697403, -0.010344876907765865, 0.05285347253084183, -0.019133444875478745, 0.0009240584331564605, 0.006114615593105555, 0...
a978db4e65cab6095f95216c7e3ced9f4946eb66
subsection
4
25
Body
An optimal distributed control problem for the two-dimensional nonlocal Cahn-Hilliard-Navier-Stokes systems with degenerate mobility and singular potential is studied in . A distributed optimal control problem for the nonlocal Cahn-Hilliard-Navier-Stokes system with non-constant viscosity and regular potential is exami...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ 0.007280521094799042, -0.026788046583533287, -0.01697898469865322, 0.010922688990831375, -0.0031902336049824953, 0.03243245184421539, -0.0008585790055803955, -0.0003758964885491878, 0.024026865139603615, 0.056627124547958374, -0.030830662697553635, 0.0003646934637799859, -0.00490453140810132...
759f13ed9de3b82072023567822c499b73f16ba0
subsection
5
25
Body
In the system (REF )-(), \mathrm {U} is the distributed control acting on the system.Let us introduce the following functional spaces required for getting the unique global solvability results of the system (REF )-().Let us denote \Vert \cdot \Vert and (\cdot , \cdot ) the norm and the scalar product, respectively, on ...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.06408021599054337, -0.0016229839529842138, -0.038539670407772064, 0.0006641647196374834, 0.008414342068135738, 0.006472864653915167, -0.013289015740156174, 0.0208870992064476, 0.022290760651230812, 0.03634263575077057, -0.028622495010495186, 0.023297734558582306, 0.009680689312517643, 0...
bf9c2e5cf8f4b00a4dbcbc7afa97f144a5e5f79b
subsection
6
25
Body
We know that \mathbf {u} can be expressed as \mathbf {u}=\sum \limits _{j=1}^{\infty }\langle \mathbf {u},\mathbf {e}_j\rangle \mathbf {e}_j, so that \mathrm {A}\mathbf {u}=\sum \limits _{j=1}^{\infty }\lambda _j\langle \mathbf {u},\mathbf {e}_j\rangle \mathbf {e}_j. Thus, it is immediate thatwhich is the Poincaré ineq...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.0748710110783577, 0.0003308404702693224, -0.08323068916797638, -0.023111160844564438, 0.03609307482838631, 0.048113930970430374, 0.009686856530606747, 0.031226763501763344, 0.010030091740190983, 0.03685582056641579, -0.04225604981184006, 0.0008170901564881206, 0.0322488434612751, -0.035...
6a5c839fc77a7ff4a44ba866ee683009f891da4d
subsection
7
25
Body
Then, if 0< \alpha < 1 and \frac{n}{2} = \alpha s_1 + (1-\alpha )s_2, the following inequality holdsFor \mathbf {u}\in \mathbb {H}^2(\Omega )\cap \mathbb {H}_0^1(\Omega ), the Agmon's inequality in 2D states that there exists a constant C>0 such thatFor every f \in \mathrm {V}^{\prime } we denote \overline{f} the avera...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.029976287856698036, -0.013935541734099388, -0.027596490457654, 0.020045211538672447, 0.013516025617718697, 0.0018191717099398375, 0.026452356949448586, 0.02169276401400566, 0.03499521687626839, 0.001042114570736885, -0.04405675455927849, -0.012158321216702461, 0.009603090584278107, 0.00...
2dca79807e2a32877188378a33f4e44327fdab09
subsection
8
25
Body
Let us first make the following assumptions:Assumption Let \mathrm {J} and \mathrm {F} satisfy:\mathrm {J}\in \mathrm {W}^{1,1}(\mathbb {R}^2;\mathbb {R}), \ \mathrm {J}(x)= \mathrm {J}(-x) \; \text{and} \ a(x) = \int _\Omega \mathrm {J}(x-y)\mathrm {d}y \ge 0, a.e., in \Omega . \mathrm {F}\in \mathrm {C}^{2}(\mathbb ...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.01427211333066225, 0.04164924472570419, -0.046683769673109055, 0.022609589621424675, -0.007147498894482851, -0.033594004809856415, 0.009344382211565971, 0.04186283051967621, -0.000031108211260288954, -0.0007513645687140524, 0.00918419286608696, -0.04528020694851875, -0.011259026825428009,...
5b26bcf6b84fa614b38ce74620dc4655350a79e1
subsection
9
25
Body
Indeed, it can be replaced by \mathrm {J}\in \mathrm {W}^{1,1}(\mathrm {B}_\delta ;\mathbb {R}), where \mathrm {B}_\delta := \lbrace z \in \mathbb {R}^2 : |z| < \delta \rbrace with \delta := \emph {diam}(\Omega )=\sup \limits _{x,y\in \Omega }d(x,y), where d(\cdot ,\cdot ) is the Euclidean metric on \mathbb {R}^2, or a...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.05280432850122452, 0.021152254194021225, -0.04242659732699394, -0.019580332562327385, -0.01893935538828373, -0.04691343754529953, 0.024296095594763756, 0.06083180382847786, -0.01190386526286602, -0.00541778514161706, -0.026051152497529984, 0.0033784848637878895, -0.0132773881778121, 0.0...
21e59bb9a035b25fb10538bc1003d4c0ee8704c8
subsection
10
25
Body
Then (\mathbf {u},\varphi ) is said to be a weak solution to the uncontrolled system (REF )-() on [0,T] corresponding to the initial conditions \mathbf {u}_0 and \varphi _0 if(i) \mathbf {u},\varphi and \mu satisfy \mathopen {}\mathclose {\left\lbrace \begin{aligned}& \mathbf {u}\in \mathrm {L}^{\infty }(0,T;\mathbb ...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ 0.004951984155923128, 0.029482999816536903, -0.014115825295448303, -0.02931513637304306, -0.021181367337703705, -0.024782810360193253, 0.05557820200920105, 0.0827721506357193, 0.0016700547421351075, 0.0591491237282753, -0.005951537285000086, -0.013635124079883099, 0.025408485904335976, 0.0...
afd6e9d286c9795a1befca08a2a31f26c758b034
subsection
11
25
Body
(iii) Moreover, the following initial conditions hold in the weak sense \mathbf {u}(0)=\mathbf {u}_0,\ \varphi (0)=\varphi _{0}, i.e., for every \mathbf {v}\in \mathbb {V}_{\text{div}}, we have (\mathbf {u}(t),\mathbf {v}) \rightarrow (\mathbf {u}_0,\mathbf {v}) as t\rightarrow 0, and for every \chi \in \mathrm {V}, ...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.016293879598379135, 0.02062670886516571, -0.03725622966885567, -0.007910464890301228, 0.003909463062882423, -0.02265581488609314, 0.06358885020017624, 0.08128631860017776, 0.014730094000697136, 0.03335057944059372, 0.0004128774453420192, -0.00858937669545412, -0.006617481354624033, 0.01...
21099e29435b1b448ce838ed968ba5237368b142
subsection
12
25
Body
Furthermore, settingthe following energy estimate holds for almost any t>0:or the weak solution (\mathbf {u},\varphi ) satisfies the following energy identity,Furthermore, if in addition \mathbf {h} \in \mathrm {L}^2_{\text{tb}}([0,\infty );\mathbb {V}_{\text{div}}^{\prime }), then the following dissipative estimate is...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.046526942402124405, 0.02425502985715866, -0.03792327269911766, -0.016475114971399307, -0.02152443118393421, -0.05259832739830017, 0.055008575320243835, 0.025002511218190193, 0.011517325416207314, 0.016017472371459007, 0.009656247682869434, -0.01575814187526703, -0.013462304137647152, 0....
3eb138d555310ed60677d0abd371bc89dc3ccf9c
subsection
13
25
Body
Then the following continuous dependence estimate holds:for all t \in [0,T], where \Lambda _0(t), \Lambda _1(t) and \Lambda _2(t) are continuous functions which depend on the norms of the two solutions. The functions \mathbb {Q} and \Lambda _i(t) also depend on \mathrm {F}, \mathrm {J} and \Omega .The following theorem...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.045850109308958054, 0.027595536783337593, -0.04447643458843231, 0.0010846282821148634, -0.0009844646556302905, -0.04817008599638939, 0.0231692623347044, 0.0349828377366066, 0.0007388254744000733, 0.028725000098347664, -0.0015873538795858622, -0.010294905863702297, -0.006471519824117422, ...
4b5b1bbaa6252815e0605841ddf1e93df44cb2eb
subsection
14
25
Body
Then, \mathbf {u}_1 = \mathbf {u}_2 and \varphi _1=\varphi _2, satisfying the following differential inequality:where \mathbf {u}=\mathbf {u}_1-\mathbf {u}_2, \varphi =\varphi _1-\varphi _2 and the function \Pi is given byTheorem 0.5 (Lemma 2.6, ) Let the Assumption REF be satisfied. Let \mathbf {u}_0\in \mathbb {G}_{...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.02054215595126152, 0.013994916342198849, -0.04377036169171333, -0.025074860081076622, 0.008203127421438694, 0.000338140525855124, 0.027104657143354416, 0.07704071700572968, 0.004517443012446165, 0.05530815199017525, -0.029851751402020454, 0.005890989676117897, -0.010759450495243073, 0.0...
f62fc7ca0bb07f30b9abd5181166db0b388a7bf2
subsection
15
25
Body
Since the pressure is also an unknown quantity, in order to linearize, we substitute \mathbf {u}= \mathbf {w}+ \widehat{\mathbf {u}}, =\widetilde{}+\widehat{} and \varphi = \psi + \widehat{\varphi } in (REF ) and (REF ) to getwhere \widetilde{}_\mathbf {w}= \widetilde{}-(\mathrm {F}^{\prime }(\widehat{\varphi })+a\wide...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ 0.010609597899019718, 0.0293346606194973, -0.027885470539331436, -0.009190917015075684, 0.00871039554476738, -0.02277516759932041, 0.018519125878810883, 0.0551607571542263, -0.012463035993278027, 0.05860830470919609, -0.03352968394756317, -0.017375027760863304, -0.016307203099131584, 0.005...
48abf92338fcd48b1aa41043c1af7f0d322d7ef4
subsection
16
25
Body
The associated cost functional is defined bywhere \mathbf {u}_d(\cdot )\in \mathrm {L}^2(0,T;\mathbb {V}_{\text{div}}) and \varphi _d(\cdot )\in \mathrm {L}^2(0,T;\mathrm {V}) are the desired states. Note that the cost functional is the sum of total energy and total effort by control.Let us assume thatand the initial d...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.019535934552550316, -0.019154373556375504, -0.04212435707449913, -0.008249353617429733, -0.008890376426279545, 0.016636069864034653, 0.011752085760235786, 0.010515827685594559, 0.01706341840326786, 0.05604371055960655, -0.008424871601164341, -0.006135504227131605, 0.02251211181282997, 0...
ce639b2aaf189cd57526159bf114f515b9936705
subsection
17
25
Body
Then the system (REF )-() can be written asWe define the augmented cost functional \widetilde{\mathcal {J}} bywhere \mathbf {p} and \eta denote the adjoint variables corresponding to \mathbf {u} and \varphi respectively. Corresponding to in the system (REF ), we have q in the adjoint system.Before establishing the Pont...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.038200657814741135, 0.025965463370084763, -0.03591227903962135, 0.02253289520740509, -0.002303633838891983, -0.011197797022759914, 0.027399513870477676, 0.02619430050253868, 0.000013587245121016167, 0.022471873089671135, 0.01862739771604538, 0.01899353787302971, -0.03801758587360382, 0....
a53b71f202cabfc7fa6005920679caa3f971b057
subsection
18
25
Body
Then, there exists a unique weak solution of the system (REF ) satisfyingand for all \mathbf {v}\in \mathbb {V} and \zeta \in \mathrm {H} and for almost all t \in (0,T), we havewhere \mathbf {p}(T)=\mathbf {p}_T\in \mathbb {G}_{\text{div}}, \eta (T)=\eta _T\in \mathrm {V} are satisfied in the weak sense.Using Galerkin ...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.04146832227706909, 0.013601548038423061, -0.04534358158707619, -0.013929571956396103, -0.00028964318335056305, -0.006091327406466007, 0.04931038245558739, 0.02178688906133175, -0.007121169473975897, 0.03249724954366684, -0.018094712868332863, -0.0005015713977627456, 0.023556692525744438, ...
d6b95157a542b1c25deebbc0dd21202f7e886b30
subsection
19
25
Body
From the definition of \mathcal {J}(\cdot ,\cdot ,\cdot ), this impliesSince \mathbf {u},\mathbf {u}_d \in \mathrm {L}^{2}(0,T;\mathbb {G}_{\text{div}}) and \varphi ,\varphi _d \in \mathrm {L}^{2}(0,T;\mathrm {H}), from the above relation, it is clear that, there exist a K>0, large enough such thatIn particular, there ...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.03439965099096298, 0.03885604068636894, -0.03034006617963314, -0.038581330329179764, -0.010324474424123764, -0.0023579176049679518, 0.006264887750148773, 0.019382234662771225, -0.007432400714606047, 0.03259878233075142, -0.02441856451332569, 0.0009309577872045338, 0.016207821667194366, ...
8aa638deb7896a85db32839137b71abf460e05c9
subsection
20
25
Body
Since \mathbf {u}^*\in \mathrm {C}([0,T];\mathbb {V}_{\text{div}}) and \mathrm {J}\in \mathrm {W}^{2,1}(\mathbb {R}^2;\mathbb {R}), we know that \varphi ^*\in \mathrm {C}([0,T];\mathrm {H}^2) and hence we haveHence (\mathbf {u}^*,\varphi ^*) is a unique strong solution of (REF )-() with control \mathrm {U}^*\in \mathrm...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.05226123705506325, 0.012218200601637363, -0.05226123705506325, -0.023032186552882195, 0.015995845198631287, 0.008036080747842789, -0.005674098618328571, 0.034800123423337936, -0.018789013847708702, 0.06471601873636246, -0.00528870290145278, -0.005323044955730438, 0.035135913640260696, 0...
60fa87dfe627a25a7b437ec5debe50da43c6740b
subsection
21
25
Body
Equivalently the above minimum principle may be written in terms of the Hamiltonian formulation. Let us first define the Lagrangian byThen, we can define the corresponding Hamiltonian bywhere {N}_1 and {N}_2 are defined by (REF ). Hence, we get the minimum principle asfor all \mathrm {W} \in \mathbb {G}_{\text{div}} an...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.046847276389598846, -0.0028459338936954737, -0.01384054683148861, -0.04144534096121788, 0.00995695311576128, -0.04300183057785034, 0.029130764305591583, -0.009262637235224247, 0.01866261102259159, 0.032320041209459305, -0.03296094760298729, 0.007149168755859137, 0.0062564765103161335, 0...
abb45de18c1323f5b56d6d4fbaf6ce88c123162f
subsection
22
25
Body
Let \mathrm {U}^*+\lambda \mathrm {U}\in \mathbb {G}_{\text{div}} such that (\mathbf {u}_{\mathbf {u}^*+\lambda \mathrm {U}},\varphi _{\mathbf {u}^*+\lambda \mathrm {U}},\mathrm {U}_{\mathbf {u}^*+\lambda \mathrm {U}})\in {A}_{\text{ad}}, for all 0\le \lambda \le 1. Then, for \lambda \in [0,1], we can deduceSince (\mat...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.039073724299669266, 0.020941073074936867, -0.008577903732657433, -0.026328973472118378, 0.0038234249223023653, -0.015003698877990246, -0.0033197400625795126, 0.04459899291396141, 0.02918318659067154, 0.030068451538681984, -0.014293961226940155, 0.016255280002951622, -0.025001076981425285,...
454ee8842a382271e719900fd659af1fd5799f31
subsection
23
25
Body
Let (\mathbf {w},\psi ) satisfy the linearized system (REF )-() with control \mathrm {U}, and initial data and forcing term to be equal to zero, that is, \mathbf {w}(0)=\widetilde{\mathbf {h}}=\mathbf {0} and \psi (0)=0. From Lemma REF (see below), we haveandsince \mathbf {u}_{\mathrm {U}^*}-\mathbf {u}_d\in \mathrm {L...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.03821032494306564, 0.015366531908512115, 0.004101048223674297, -0.013130984269082546, -0.0004730511282104999, -0.0021611570846289396, 0.00798846036195755, 0.03640967607498169, -0.005859730299562216, 0.04699992015957832, -0.00629081716760993, 0.007618412375450134, 0.021424638107419014, -...
0803a39d4ad58163ab9b463b1fbb978c59dff2eb
subsection
24
25
Body
Thus it is immediate thatSince the above equality is true for all \mathrm {U}\in \mathbb {G}_{\text{div}}, we getLemma Let (\mathbf {u}_0, \varphi _0) satisfies (REF ) and \mathrm {F}(\cdot ) satisfies REF , the mapping \mathrm {U}\mapsto (\mathbf {u}_{\mathrm {U}},\varphi _{\mathrm {U}}) from {U}_{\text{ad}} into \ma...
{ "cite_spans": [] }
1802.08413
Pontryagin's maximum principle and second order optimality condition for optimal control problems for the nonlocal Cahn-Hilliard-Navier-Stokes systems in two dimensions
[ "Tania Biswas", "Sheetal Dharmatti", "Manil T Mohan" ]
[ "math.OC" ]
2,018
en
Mathematics
[ -0.015987740829586983, 0.035667307674884796, -0.014035039581358433, -0.032646723091602325, -0.0060526104643940926, -0.02634621039032936, -0.010411863215267658, 0.03728438541293144, 0.009405001997947693, 0.02598007768392563, -0.0030472816433757544, -0.0032913691829890013, 0.001724822446703910...
02b9c7614332f10d30f9bb2b0611a6d762de3c9c
abstract
0
59
Abstract
Completion is one of the most studied techniques in term rewriting and fundamental to automated reasoning with equalities. In this paper we present new correctness proofs of abstract completion, both for finite and infinite runs. For the special case of ground completion we present a new proof based on random descent. ...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.028061101213097572, 0.030670372769236565, 0.0009226861293427646, 0.022308500483632088, -0.00024891068460419774, -0.05654945224523544, 0.01760876178741455, 0.004688294604420662, 0.030761925503611565, 0.007190752774477005, -0.055389776825904846, -0.018432741984725, 0.033539045602083206, 0...
acd03cddde230b7c2c43727b92bdfc7ad5b79806
subsection
1
59
Introduction
Reasoning with equalities is pervasive in computer science and mathematics, and has consequently been one of the main research areas of automated deduction. Indeed completion as introduced by Knuth and Bendix  has evolved into a fundamental technique whose ideas appear throughout automated reasoning whenever equalities...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.010411947965621948, 0.029595866799354553, -0.01621670462191105, 0.019679725170135498, -0.02212062105536461, -0.03258596360683441, 0.031487561762332916, -0.006033590063452721, 0.03655242174863815, 0.01739138551056385, -0.03524043783545494, -0.010923010297119617, 0.016765905544161797, 0.0...
5267acfa910b8914c6f2ed02876154873090172d
subsection
2
59
Introduction
In this paper we revisit and extend his work.A special case of KB_\mathsf {f} that is known to be decidable is the completion of ground systems . We present new correctness and completeness proofs for the corresponding inference system KB_\mathsf {g}, based on the recent notion of random descent .On a given set of inpu...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.02664361149072647, 0.046969667077064514, -0.025087112560868263, 0.001968514174222946, -0.008881203830242157, -0.045870959758758545, 0.05252423509955406, -0.023988407105207443, 0.022004632279276848, 0.052402153611183167, -0.018891632556915283, -0.01169663667678833, 0.005562197417020798, ...
b3016a3bd68dfd9f56c000f387e515f6c0bb41d2
subsection
3
59
Introduction
It moreover incorporated canonicity results (Section ). In addition to these results we present new and formalized proofs of correctness and completeness of ground completion (Section ), as well as completeness of ordered completion for two different cases (Section ). At the end of each section, we remark on the novelt...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ 0.014446365647017956, 0.03432346507906914, -0.010907234624028206, 0.03346919268369675, -0.011097921058535576, -0.056565072387456894, 0.014499757438898087, 0.005457431077957153, 0.016109146177768707, 0.03578793257474899, -0.032981038093566895, 0.018931297585368156, 0.0060142339207232, 0.024...
a6311a319c36caa007bd235a9514fc41f2eda972
subsection
4
59
Preliminaries
We assume familiarity with the basic notions of abstract rewrite systems, term rewrite systems, and completion , , but nevertheless shortly recapitulate terminology and notation that we use in the remainder.
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.021959353238344193, 0.027605608105659485, -0.002212721388787031, 0.06641216576099396, -0.02417207509279251, -0.08399185538291931, 0.0299556702375412, 0.019731370732188225, 0.0021230680868029594, 0.00004631693445844576, 0.0003254703478887677, -0.02826179377734661, -0.037356842309236526, ...
76b7f33eb9d9496ba993483fb408004917066043
subsection
5
59
Rewrite Systems
For an arbitrary binary relation \mathchoice{\xrightarrow[\alpha ]{}}{\rightarrow {\alpha }{}{_{\alpha }}{}{}{^{}}}{\rightarrow {\alpha }{}{_{\alpha }}{}{}{^{}}}{\rightarrow {\alpha }{}{_{\alpha }}{}{}{^{}}}, we write \mathchoice{\xleftarrow[\alpha ]{}}{\textstyle \vphantom{\leftarrow }{\alpha }{}{_{\vphantom{\alpha }}...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.049182258546352386, -0.011876051314175129, -0.002431272529065609, 0.02730652689933777, -0.0030452879145741463, -0.01855013146996498, 0.029610037803649902, 0.0473516546189785, 0.01775687001645565, -0.010960748419165611, 0.0012242172379046679, 0.004919751547276974, 0.002467503072693944, -...
259c0d24ea41dd868a4fb4ec02d2163b8a345e9f
subsection
6
59
Rewrite Systems
We further use as abbreviation for the \emph {joinability relation} []******* []*\textstyle \vphantom{\leftarrow }{\alpha }{}{_{\vphantom{\alpha }}}{*}{}{^{*}}\textstyle \vphantom{\leftarrow }{\alpha }{}{_{\alpha }}{*}{}{^{\vphantom{*}}}\scriptstyle \vphantom{\leftarrow }{\alpha }{}{_{\vphantom{\alpha }}}{*}{}{^{*}}\sc...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.029517093673348427, -0.005277991760522127, -0.03407813236117363, 0.019678061828017235, -0.004118664190173149, -0.06171894818544388, 0.028174715116620064, 0.02086789906024933, 0.04975956678390503, 0.014758546836674213, -0.050064653158187866, 0.010914459824562073, 0.009175468236207962, 0....
68af364ae1d12c46050facc2efc6cc72be03e3b6
subsection
7
59
Rewrite Systems
The subset consisting of the positions addressing function symbols in t is denoted by \mathcal {P}\mathsf {os}_\mathcal {F}(t) whereas \mathcal {P}\mathsf {os}_\mathcal {V}(t) = \mathcal {P}\mathsf {os}(t) - \mathcal {P}\mathsf {os}_\mathcal {F}(t) is the set of variable positions in t. We write p \leqslant q if p is a...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.03621398285031319, 0.008969596587121487, -0.035603806376457214, 0.05476335063576698, 0.014781529083848, -0.053939614444971085, 0.022942641749978065, 0.04924125224351883, 0.02736642211675644, 0.02846474014222622, 0.008420437574386597, 0.013980671763420105, 0.014339150860905647, 0.0346275...
5be121cc7dc334615d852ca4dabf0bb1dcc89092
subsection
8
59
Rewrite Systems
Two TRSs \mathcal {R}_1 and \mathcal {R}_2 are called literally similar, denoted by \mathcal {R}_1 \doteq \mathcal {R}_2, if every rewrite rule in \mathcal {R}_1 has a variant in \mathcal {R}_2 and vice versa. The following result is folklore; we formalized the non-trivial proof.Two terms s and t are variants of each o...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.018177548423409462, 0.027212904766201973, -0.03305841237306595, 0.031776368618011475, -0.009577171877026558, -0.04191061854362488, 0.008981937542557716, -0.003790804883465171, 0.04703879356384277, 0.042948465794324875, -0.03971283137798309, 0.013835388235747814, 0.004254401195794344, 0....
59da412df7c76f6c9f69318ebdd4bb56e6b63247
subsection
9
59
Rewrite Systems
\hfill {\normalfont \href {http://cl-informatik.uibk.ac.at/isafor/v2.37/LMCS2019/Abstract_Completion.html\#lem:SN_encomp_Un_less_relto_encompeq}{{linkblue}{{check-square-o}}}}First we show the inclusion {\mathrel {\mbox{[}0pt]{\mbox{[}9pt][r]{\cdot }}}{\unrhd }}\cdot \mathrel {R} R \cdot . Suppose s \cdot t R u. So s =...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04237886890769005, 0.032340966165065765, -0.03298168256878853, 0.012966896407306194, -0.01990799978375435, -0.02552190236747265, 0.01516364049166441, -0.011677834205329418, 0.034354645758867264, 0.03362239897251129, -0.027001654729247093, 0.008733585476875305, 0.010976096615195274, 0.03...
764199e13b8e169759dcc0c5f9fd748c357ae46b
subsection
10
59
Abstract Confluence Criteria
We use the following simple confluence criterion for ARSs to replace Newman's Lemma in the correctness proof of abstract completion. In the sequel, we will refer to a conversion of the form \mathrel {{\vphantom{\rightarrow }_{\mathcal {A}}}{\mathrel {\leftarrow }}} \cdot \rightarrow _\mathcal {A} as a peak.[Peak Decrea...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.00804161187261343, 0.03735458478331566, -0.06866651773452759, 0.011627529747784138, -0.022614171728491783, -0.06585881859064102, 0.03118985705077648, -0.009521757252514362, 0.0424511656165123, 0.045350417494773865, -0.03894154354929924, -0.00007552124588983133, -0.012306565418839455, 0....
97c96d1cfe5c09380347604a7d9f1362a840063d
subsection
11
59
Abstract Confluence Criteria
We denote by \mathcal {M}(J) the set of all multisets over a set J.Every peak decreasing ARS is confluent. linkbluecheck-square-o Let > be a well-founded order on I which shows that the ARS \mathcal {A}= \langle A, {\lbrace {\mathchoice{\xrightarrow[\alpha ]{}}{\rightarrow {\alpha }{}{_{\alpha }}{}{}{^{}}}{\rightarro...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.001250954926945269, 0.00003212007868569344, -0.08427164703607559, 0.02953474223613739, -0.021922223269939423, -0.054340261965990067, -0.016353948041796684, -0.0013577437493950129, 0.02064075693488121, 0.029794085770845413, -0.03405038267374039, 0.017376068979501724, 0.008565990254282951, ...
4ba577ec0b4efb2cbc5756ae633dca746f94e07b
subsection
12
59
Abstract Confluence Criteria
Here an ARS \mathcal {A} is called Church-Rosser modulo an ARS \mathcal {B} if the inclusion{[\mathcal {A}\hspace{0.85358pt}\cup \hspace{0.85358pt}\mathcal {B}\,]{*}} \subseteq {\xrightarrow[\mathcal {A}]{*} \cdot [\mathcal {B}]{*} \cdot \xleftarrow[\mathcal {A}]{*}}holds.[Peak Decreasingness Modulo linkbluecheck-squar...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.006319943815469742, -0.011274443939328194, -0.07976020872592926, 0.03716447204351425, -0.01850595511496067, -0.026576565578579903, 0.013547640293836594, 0.0060872845351696014, 0.037530627101659775, 0.053458258509635925, -0.028529377654194832, 0.014508790336549282, 0.010175987146794796, ...
54a23317ab46c61a3f6140dac0cc3290f219b6f6
subsection
13
59
Abstract Confluence Criteria
If the given conversion is not of the desired shape, there is an index 1 i < n such that xi xi+1 xi+2 or xi xi+1 xi+2 for some I and I J. As the reasoning is similar, we only consider the former case. By peak decreasingness there are labels 1,...,m with xi []1 []m xi+2 such that j for all 1 j m. Writing N for the multi...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.051866885274648666, 0.012440425343811512, -0.07432219386100769, 0.034659285098314285, -0.011273420415818691, -0.007902072742581367, 0.008390231989324093, 0.007574090734124184, 0.027291135862469673, 0.003655471373349428, -0.036886509507894516, -0.0043476652354002, 0.004374361597001553, 0...
5f1e06011c1a2dea7534095fe0667449ade46769
subsection
14
59
Critical Peaks
Completion is based on critical pair analysis. In this subsection we present a version of the critical pair lemma that incorporates primality (cf. Definition REF below).[Overlaps linkbluecheck-square-o] An overlap of a TRS \mathcal {R} is a triple \langle \ell _1 \rightarrow r_1, p, \ell _2 \rightarrow r_2 \rangle , co...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.036222781985998154, -0.00444392766803503, -0.010848524048924446, 0.030699340626597404, 0.004878784529864788, 0.008834452368319035, 0.035704005509614944, 0.0013102912344038486, -0.014571505598723888, 0.006843267474323511, -0.06353481858968735, 0.010322119109332561, 0.001213020645081997, ...
929fed5934acd0c03a68bdd175526a0ee91cbd0e
subsection
15
59
Critical Peaks
The term \ell _2\sigma {[\ell _1\sigma ]}_p = \ell _2\sigma can be reduced in two different ways:[minimum height=6mm] s)\ell _2\sigma [\ell _1\sigma ]_p = \ell _2\sigma ; (swest) at (s.west); (seast) at (s.east); t)[below left of=swest, anchor=north east]\ell _2\sigma [r_1\sigma ]_p; u)[below right of=seast, anchor=no...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04128022864460945, 0.02744387835264206, -0.02764219418168068, -0.004549825564026833, 0.001191802672110498, -0.03255432844161987, 0.011105693876743317, 0.03083050437271595, -0.01601782813668251, 0.021814756095409393, -0.028404947370290756, 0.0467720553278923, -0.01760435476899147, 0.0076...
24517643658fef0cf6a4f1590ec8cebc62ad2d60
subsection
16
59
Critical Peaks
If t \mathrel {{\vphantom{\xleftarrow{}}_{\mathcal {R}}}{\xleftarrow{}}} s \xrightarrow{}_\mathcal {R}u then one of the following holds: linkbluecheck-square-ot \mathrel {\downarrow }_\mathcal {R}u, p_2 \leqslant p_1 and {t|_{p_2} \xleftarrow{} s|_{p_2} \xrightarrow{} u|_{p_2}} is an instance of a critical peak, or ...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.050134941935539246, 0.03487780690193176, -0.02636432833969593, 0.013586477376520634, -0.009741679765284061, -0.0017059382516890764, 0.06322555989027023, 0.04082809016108513, 0.012243850156664848, 0.019696960225701332, -0.017133761197328568, 0.00805576704442501, -0.029553068801760674, 0....
6ba886ca365dee8a998366bd69ba666c9fddf67f
subsection
17
59
Critical Peaks
Let \sigma ^{\prime }(x) = \sigma _1(x) for x \in \mathcal {V}\mathsf {ar}(\ell _1 \rightarrow r_1) and \sigma ^{\prime }(x) = \sigma _2(x), otherwise. The substitution \sigma ^{\prime } is a unifier of {\ell _2}|_p and \ell _1: ({\ell _2}|_p)\sigma ^{\prime } = (\ell _2\sigma _2)|_p = \ell _1\sigma _1 = \ell _1\sigma ...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.03204420581459999, 0.014084191992878914, -0.014839519746601582, -0.031494878232479095, -0.028137866407632828, -0.01505314838141203, 0.010223628021776676, 0.06494292616844177, 0.03176954388618469, -0.0016479878686368465, 0.014755594544112682, -0.0007939524948596954, -0.024002637714147568, ...
c2d13745334c830c686865b0b89518d77e62360f
subsection
18
59
Critical Peaks
We also have \ell _2\sigma _2{[r_1\sigma _1]}_p = \ell _2\sigma _2{[\sigma _2^{\prime }(x)]}_{q_1} \rightarrow ^* \ell _2\sigma _2^{\prime } \rightarrow r_2\sigma _2^{\prime } Consequently, t \rightarrow ^* s{[r_2\sigma _2^{\prime }]}_{p_2} \mathrel {{\vphantom{\rightarrow }^{*}}{\mathrel {\leftarrow }}} u. Hence, (R...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04033772647380829, 0.027110492810606956, -0.023754101246595383, -0.018704256042838097, -0.0013644876889884472, 0.003364020027220249, 0.02132834494113922, 0.023479485884308815, 0.008543542586266994, 0.0404902920126915, -0.020443476736545563, 0.03319776803255081, -0.010671800002455711, -0...
e90ff8b84f6ba840d9af3cd7ed64b9290f5780a9
subsection
19
59
Critical Peaks
Now, if \smash{v \xleftarrow{} s \xrightarrow{} u} is an instance of a critical peak then v \rightarrow _{\mathsf {PCP}(\mathcal {R})} u. Otherwise, v \mathrel {\downarrow }_\mathcal {R}u by Lemma REF , since q \lnot \leqslant \epsilon . In both cases we obtain v \mathrel {\triangledown _{\hspace{-1.42262pt}s}} u. Fina...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.024784788489341736, 0.01176666934043169, -0.03845915198326111, -0.01687929593026638, 0.014086428098380566, 0.0020164349116384983, 0.03687195107340813, 0.005707826931029558, 0.030752060934901237, 0.03699404373764992, -0.01846649870276451, 0.024418510496616364, -0.004895148333162069, 0.01...
fb83cd91bb18a567d82bbf114bac847974a7a877
subsection
20
59
Critical Peaks
Lemma REF yields a term v such that t \mathrel {\triangledown _{\hspace{-1.42262pt}s}} v \mathrel {\triangledown _{\hspace{-1.42262pt}s}} u. From the assumption \mathsf {PCP}(\mathcal {R}) \subseteq {\mathrel {\downarrow }_\mathcal {R}} we obtain t \mathrel {\downarrow }_\mathcal {R}v \mathrel {\downarrow }_\mathcal {R...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.060378555208444595, -0.001196200493723154, -0.04444447532296181, 0.010119054466485977, 0.03044874034821987, 0.038431040942668915, 0.014407824724912643, -0.010989018715918064, 0.02651100791990757, 0.012026870623230934, -0.05894387513399124, 0.023290615528821945, -0.006643014494329691, -0...
9cc6bf885e56edb2a96a15b44d13aa1846196b5b
subsection
21
59
Correctness for Finite Runs
The original completion procedure by Knuth and Bendix  was presented as a concrete algorithm. Later on, Bachmair, Dershowitz, and Hsiang  presented an inference system for completion and showed that all fair implementations thereof (in particular the original procedure) are correct. Abstracting from a concrete strategy...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.026695141568779945, 0.03711983188986778, -0.018758336082100868, 0.012477574869990349, -0.0013364742044359446, -0.04606400057673454, 0.07344598323106766, 0.009371536783874035, 0.03507457673549652, 0.03226616978645325, -0.02811460942029953, 0.0025489358231425285, -0.000944403582252562, 0....
8b524fd647eb6cb6e377808772663d303331651c
subsection
22
59
Correctness for Finite Runs
Then, the following two inclusions hold:If s \xrightarrow[\mathcal {E}\hspace{0.85358pt}\cup \hspace{0.85358pt}\mathcal {R}]{} t then s \xrightarrow[\mathcal {R}^{\prime }]{=} \cdot \xrightarrow[\mathcal {E}^{\prime } \hspace{0.85358pt}\cup \hspace{0.85358pt}\mathcal {R}^{\prime }]{=} \cdot \xleftarrow[\mathcal {R}^{\p...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04199666902422905, 0.05258740112185478, -0.011575027368962765, 0.023012710735201836, 0.011300339363515377, -0.018419325351715088, -0.017152709886431694, 0.027407711371779442, 0.048436567187309265, 0.04214927554130554, -0.006191914435476065, 0.004322513472288847, -0.01077385526150465, 0....
ea6583a11372ac78f17a8ca01a72c954149f9aad
subsection
23
59
Correctness for Finite Runs
If s \approx t \in \mathcal {E} then s \approx t \in \mathcal {E}^{\prime } because \mathcal {E}\subseteq \mathcal {E}^{\prime }. If s \approx t \in \mathcal {R} then either s \approx t \in \mathcal {R}^{\prime } or s \rightarrow _\mathcal {R}u with u \approx t \in \mathcal {E}^{\prime } and thus s \rightarrow _{\mathc...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04981257766485214, 0.0221592728048563, -0.0027164947241544724, 0.014185291714966297, -0.00879045482724905, -0.021152032539248466, 0.017229903489351273, 0.02859950251877308, 0.04138839244842529, 0.0443490669131279, -0.01384191494435072, 0.019610650837421417, -0.015390926972031593, 0.0551...
cf25159fe91d91a568b192aa4ef57c2c701bf068
subsection
24
59
Correctness for Finite Runs
It is the final result in this section whose proof refers to the inference rules.If (\mathcal {E},\mathcal {R}) \vdash _\mathsf {f}^* (\mathcal {E}^{\prime },\mathcal {R}^{\prime }) and \mathcal {R}\subseteq {>} then \mathcal {R}^{\prime } \subseteq {>}. linkbluecheck-square-o We consider a single step (\mathcal {E},...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.029347211122512817, 0.027149604633450508, 0.015093287453055382, 0.014398904517292976, -0.016008958220481873, -0.005196425132453442, 0.006642420310527086, 0.027805835008621216, 0.029331950470805168, 0.05481809005141258, -0.023593753576278687, 0.0187254436314106, 0.014040268026292324, 0.0...
f6a7bc0b74872623ff4be1c54e3b9e506a6e0ab3
subsection
25
59
Correctness for Finite Runs
The run is fair if \mathcal {E}_n= \varnothing and\mathsf {PCP}({\mathcal {R}_n}) ~\subseteq ~ {\mathrel {\downarrow }_{{\mathcal {R}_n}}} \cup \bigcup _{i=0}^n {[\mathcal {E}_i]{}}The reason for writing []{}_{\mathcal {E}_i} instead of \mathcal {E}_i in the definition of fairness is that critical pairs are ordered, so...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.042668987065553665, 0.014131431467831135, -0.018175523728132248, 0.0305672325193882, -0.006123365834355354, 0.0325053445994854, -0.0477660670876503, 0.029453199356794357, 0.05197802931070328, 0.02882750891149044, -0.042668987065553665, -0.006756685674190521, 0.007576949894428253, 0.0299...
71c73fd0e3140dedfb5b495bb6502b58e48db83f
subsection
26
59
Correctness for Finite Runs
If \smash{t \mathrel {\smash{[\mathcal {E}\hspace{0.85358pt}\cup \hspace{0.85358pt}\mathcal {R}]{M}}^{*}\vphantom{[\mathcal {E}\hspace{0.85358pt}\cup \hspace{0.85358pt}\mathcal {R}]{M}}} u} and \mathcal {R}^{\prime } \subseteq {>} then \smash{t \mathrel {\smash{[\mathcal {E}^{\prime } \hspace{0.85358pt}\cup \hspace{0.8...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.024056684225797653, 0.02956363558769226, 0.004050126299262047, -0.016993751749396324, 0.017832759767770767, 0.03886901214718819, -0.019800618290901184, 0.00329692498780787, 0.02501773089170456, 0.026421165093779564, 0.025063494220376015, 0.009313003160059452, 0.008100252598524094, 0.004...
daa92c4b42646d0fe73434042c83a7b6f385fe5a
subsection
27
59
Correctness for Finite Runs
Hencet \mathrel {\smash{\xrightarrow[\mathcal {R}^{\prime }]{M}}^{=}\vphantom{\xrightarrow[\mathcal {R}^{\prime }]{M}}} v \mathrel {\smash{\xrightarrow[\mathcal {E}^{\prime } \hspace{0.85358pt}\cup \hspace{0.85358pt}\mathcal {R}^{\prime }]{M}}^{=}\vphantom{\xrightarrow[\mathcal {E}^{\prime } \hspace{0.85358pt}\cup \hsp...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.021215317770838737, 0.04737578704953194, 0.008379287086427212, -0.0477115698158741, 0.004040831234306097, 0.02037586271762848, -0.023611579090356827, -0.0050634401850402355, 0.010149774141609669, 0.02072690799832344, -0.02678624540567398, 0.032143495976924896, 0.012011838145554066, 0.00...
850f839371bec92d2542c0c456be3267846c6497
subsection
28
59
Correctness for Finite Runs
Because s > v and s > w we have M_1 >_{\mathsf {mul}}\lbrace v, w \rbrace and M_2 >_{\mathsf {mul}}\lbrace v, w \rbrace . Hence by repeated applications of Lemma  we obtain a conversion in \mathcal {R}_n between v and w in which each step is labeled with a multiset that is smaller than both M_1 and M_2. It follows that...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.0238033477216959, 0.037230879068374634, -0.0160672590136528, 0.00939164124429226, -0.023421883583068848, 0.028136778622865677, 0.00396722462028265, -0.008262508548796177, 0.04437188059091568, 0.02249111235141754, -0.04138120636343956, 0.0313410758972168, 0.004356317687779665, 0.02673299...
f6b3d293e8cce4cb07912d89427d3a2a71f8ace2
subsection
29
59
Correctness for Finite Runs
One possible run is(\mathcal {E},\varnothing ) ~ & \mathrel {{\vdash _\mathsf {f}^\textsf {\small \tiny orient}}^+} && (\lbrace \mathsf {a} \approx \mathsf {c}, \mathsf {f}(\mathsf {a}) \approx \mathsf {d} \rbrace , \lbrace \mathsf {a} \rightarrow \mathsf {b},\mathsf {f}(\mathsf {b}) \rightarrow \mathsf {b} \rbrace ) \...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.019064050167798996, 0.05424630269408226, -0.02378045581281185, 0.03602174296975136, -0.036662809550762177, 0.014744493179023266, -0.013172358274459839, 0.05620002746582031, 0.019888276234269142, 0.031595341861248016, -0.010134931653738022, 0.028924239799380302, -0.003375130472704768, -0...
65e3f8452703b0741d90f9f060ce5aace814a579
subsection
30
59
Correctness for Finite Runs
However, the run(\mathcal {E},\varnothing ) ~ & \vdash _\mathsf {f}^\textsf {\small \tiny orient} && (\lbrace \mathsf {a} \approx \mathsf {b}, \mathsf {f}(\mathsf {b}) \approx \mathsf {b}, \mathsf {f}(\mathsf {a}) \approx \mathsf {d} \rbrace , \lbrace \mathsf {a} \rightarrow \mathsf {c} \rbrace ) \\ & \mathrel {{\vdash...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.03246375918388367, 0.06730738282203674, -0.013630812987685204, 0.005781844723969698, -0.03093820810317993, -0.012929058633744717, 0.002635392127558589, 0.0037395108956843615, 0.011006861925125122, 0.0249275304377079, -0.019649118185043335, 0.021342482417821884, 0.004672004841268063, 0.0...
5de226b5e697c8c381f4707c9cff19ba275f2c65
subsection
31
59
Correctness for Finite Runs
When applying compose steps in a naive way by simplifying the rules in descending order, exponentially many steps are required to obtain a canonical system . However, when processing the rules in reverse order only a polynomial number of steps is necessary.This section resumes our results on finite runs . The presented...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ 0.0010469711851328611, 0.012388205155730247, -0.038263075053691864, 0.050376664847135544, 0.0005663942429237068, -0.021175896748900414, 0.02578333392739296, 0.014859743416309357, -0.020321538671851158, 0.014813974499702454, -0.038293588906526566, -0.03338102623820305, 0.003169519128277898, ...
d5f05e464a02f0c808e10ebf8aaeca1bdfac8562
subsection
32
59
Canonicity and Normalization Equivalence
A natural question arising in the context of completion concerns uniqueness of resulting systems: Is there a single complete presentation of a given equational theory conforming to a certain reduction order? Métivier  showed that for reduced and hence canonical systems this is indeed the case, up to renaming variables....
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.01052834838628769, 0.015975624322891235, -0.06115597113966942, 0.06909038126468658, -0.00819380208849907, -0.04864402115345001, 0.0007285922183655202, 0.007942036725580692, 0.0178524162620306, 0.04864402115345001, -0.06732039898633957, -0.03771895170211792, 0.008476083166897297, 0.03964...
d72a20e7206aac66ea9ac431b88b7ae4606e37ac
subsection
33
59
Canonicity and Normalization Equivalence
However, the present version suffices to prove the following lemma that we employ in our proof of Métivier's transformation result  (Theorem  below).Let \mathcal {A} and \mathcal {B} be ARSs such that \mathsf {NF}(\mathcal {B}) \subseteq \mathsf {NF}(\mathcal {A}) and either{\rightarrow _\mathcal {B}} \subseteq {\right...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ 0.007959475740790367, 0.020711425691843033, -0.03791243955492973, 0.012011711485683918, 0.005960067734122276, -0.006742279045283794, 0.003140291664749384, 0.011653038673102856, 0.06746094673871994, 0.05955488979816437, -0.021734023466706276, -0.012538272887468338, 0.01633104309439659, 0.02...
fafaa83ab5a8f0cfddf698f82c86d5192eea4581
subsection
34
59
Canonicity and Normalization Equivalence
\cdot \mathrel {\vphantom{\mathrel {\leftarrow }}^{!}_{\mathcal {A}}}\scriptstyle !}\hspace{0.0pt}\copy 1\scriptstyle \mathcal {A}    B! \scriptstyle !\scriptstyle \mathcal {B}where we obtain the inclusions fromB A, confluence of A, termination of A, and normalization equivalence of A and B, respectively.In the above ...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.049656543880701065, 0.02490457147359848, -0.01448188629001379, 0.005611921660602093, 0.009384995326399803, -0.02760561928153038, 0.0432167574763298, 0.008713548071682453, 0.04166022315621376, 0.019853461533784866, -0.04568890109658241, 0.021547337993979454, 0.0014106105081737041, 0.0299...
ee9e12493fbf1a57488b1645da281a4047c57786
subsection
35
59
Canonicity and Normalization Equivalence
(This is the only place in the paper where variant-freeness of TRSs is important.)The following example shows why the result of \dot{\mathcal {R}} has to be variant-free.Consider the TRS \mathcal {R} consisting of the four rules\mathsf {f}(x) &\rightarrow \mathsf {a} & \mathsf {f}(y) &\rightarrow \mathsf {b} & \mathsf ...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.025600528344511986, 0.02581411972641945, -0.014379318803548813, 0.02932313084602356, 0.03213034197688103, -0.018445195630192757, 0.014943812042474747, 0.0299944207072258, 0.02132868766784668, 0.03396112844347954, -0.039941705763339996, -0.009077660739421844, 0.01117543876171112, -0.0006...
1818536f22f53255c41bf6a3fc789c4b9990ea09
subsection
36
59
Canonicity and Normalization Equivalence
\begin{} \item If \ell \mathrel {{\hss \rhd \hss \cr \hspace{1.69997pt}\scalebox {1.0}{\cdot }}}\end{}\ell ^{\prime } then we obtain \ell ^{\prime } \notin \mathsf {NF}(\ddot{\mathcal {R}}) from the induction hypothesis and hence \ell \notin \mathsf {NF}(\ddot{\mathcal {R}}) as desired.If \ell \doteq \ell ^{\prime } th...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.062757708132267, 0.028509581461548805, 0.011660235933959484, -0.004410743713378906, 0.007936285808682442, -0.027624379843473434, -0.009630377404391766, 0.04212336614727974, 0.01904708705842495, 0.041634976863861084, -0.04807557910680771, 0.014995002187788486, -0.011347362771630287, 0.03...
2e28c91d72e352949c71023ded0bccd1704d88e5
subsection
37
59
Canonicity and Normalization Equivalence
\hfill {\normalfont \href {http://cl-informatik.uibk.ac.at/isafor/v2.37/LMCS2019/Normalization_Equivalence.html\#lem:right_reduced_min_step_rule}{{linkblue}{{check-square-o}}}} Let \ell \rightarrow r be the rewrite rule that is used in the first step from s to t. So s \mathrel {\mbox{[}0pt]{\mbox{[}9pt][r]{\cdot }}}{\...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.050742991268634796, 0.03957526013255119, -0.017819546163082123, 0.0444878414273262, 0.009176760911941528, 0.012502670288085938, 0.0150733832269907, 0.012983248569071293, 0.03621883690357208, 0.041833218187093735, -0.028178684413433075, -0.009275928139686584, -0.014615689404308796, 0.018...
02c213e7efee225673f5a62dcca82bb2bf16e602
subsection
38
59
Canonicity and Normalization Equivalence
Let \mathcal {R} and \mathcal {S} be equivalent canonical TRSs. If \mathcal {R} and \mathcal {S} are compatible with the same reduction order then \mathcal {R}\doteq \mathcal {S}. linkbluecheck-square-o Suppose \mathcal {R} and \mathcal {S} are compatible with the reduction order >. We show that {\rightarrow _\mathcal ...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.027677731588482857, 0.05007624626159668, -0.02702164277434349, -0.002826515818014741, 0.03289591521024704, -0.052975237369537354, 0.019011275842785835, -0.006602832581847906, 0.04818427562713623, 0.0783642902970314, -0.01920962706208229, 0.0012063233880326152, 0.004581167828291655, 0.04...
1c2f8fa62193ec16365f90bc48c4335c00215206
subsection
39
59
Canonicity and Normalization Equivalence
Consider the ES \mathcal {E} consisting of the ground equations \mathsf {f}(\mathsf {f}(\mathsf {f}(\mathsf {a}))) &\approx \mathsf {f}(\mathsf {b}) & \mathsf {f}(\mathsf {f}(\mathsf {b})) &\approx \mathsf {c} & \mathsf {f}(\mathsf {c}) &\approx \mathsf {a} & \mathsf {f}(\mathsf {a}) &\approx \mathsf {f}(\mathsf {f}(\...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ 0.009495565667748451, 0.022743595764040947, -0.012165002524852753, -0.007695603184401989, -0.013484465889632702, -0.039873749017715454, 0.02838754653930664, 0.019402988255023956, 0.02556557208299637, 0.04927016422152519, -0.050124384462833405, 0.013553109019994736, 0.006143516860902309, 0....
b6c7db439004cea099c0ab7940db625c6c1ef429
subsection
40
59
Canonicity and Normalization Equivalence
We start by applying \textsf {\small orient} to the last two equations:} \mathsf {f}(\mathsf {f}(\mathsf {f}(\mathsf {a}))) &\approx \mathsf {f}(\mathsf {b}) & \mathsf {f}(\mathsf {f}(\mathsf {b})) &\approx \mathsf {c} & \mathsf {f}(\mathsf {c}) &\mathrel {\leftarrow }\mathsf {a} & \mathsf {f}(\mathsf {a}) &\rightarrow...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.026432327926158905, 0.061349861323833466, -0.024448378011584282, -0.0003145229711662978, -0.008256287313997746, 0.0007754576508887112, 0.002853836864233017, 0.023319052532315254, 0.038915958255529404, 0.04444049671292305, -0.029988178983330727, 0.004929354414343834, 0.027943182736635208, ...
fc22187392826076533cb2b0ff099943d9dbe2f0
subsection
41
59
Canonicity and Normalization Equivalence
The absence of deduce from KB_\mathsf {g} does not hurt for ground systems. If s \mathrel {\leftarrow }\cdot \rightarrow t and the two contracted redexes are at parallel positions then trivially s \rightarrow \cdot \mathrel {\leftarrow }t. If the steps are identical then s = t. In the remaining case one of the contract...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.061639539897441864, 0.023343687877058983, -0.0325591042637825, 0.00384674989618361, -0.022672366350889206, -0.013083145022392273, 0.03893665969371796, 0.011313296854496002, 0.026395149528980255, 0.03631240129470825, -0.026074746623635292, 0.0033508872147649527, -0.0037666489370167255, 0...
038b4a9f29fca5aa0741fe8283916b7535e9e669
subsection
42
59
Canonicity and Normalization Equivalence
Furthermore let M(\mathcal {E},\mathcal {R}) denote the (finite) multiset of left-hand sides and right-hand sides occurring in \mathcal {E} and \mathcal {R} M(\mathcal {E},\mathcal {R}) = \bigcup \,\lbrace \lbrace s, t \rbrace \mid (s,t) \in \mathcal {E}\rbrace ~\cup ~ \bigcup \,\lbrace \lbrace s, t \rbrace \mid (s,t...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.02487003244459629, 0.026273740455508232, -0.01942303776741028, 0.0100471880286932, -0.009749663062393665, 0.0052524590864777565, -0.006240394432097673, 0.008681625127792358, 0.021696433424949646, 0.03860195353627205, -0.03259042277932167, 0.016997065395116806, 0.02319168671965599, 0.033...
e03582671cbf237b9d67e812dddcfaf807014103
subsection
43
59
Canonicity and Normalization Equivalence
In particular, it holds for any LPO or KBO based on a total precedence. Next we consider completeness of ground completion. Our proof makes use of the following concept. [Random Descent linkbluecheck-square-o] An ARS \mathcal {A} has random descent if for every conversion a b with normal form b we have a \rightarrow ^n...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.01898578554391861, 0.03965038061141968, -0.021580306813120842, 0.01009573694318533, 0.003588450839743018, 0.009897332638502121, 0.04712870344519615, -0.013621232472360134, 0.01897052302956581, 0.04715922474861145, -0.05192093551158905, 0.01367464940994978, 0.017642740160226822, 0.051188...
25547b1ed848e5a00420d1ecba227118750f815c
subsection
44
59
Canonicity and Normalization Equivalence
Left-reduced TRSs enjoy the WCR1 property. linkbluecheck-square-o This follows from a straightforward case analysis on the relative positions of the two redexes that are part of a peak together with the fact that for left-reduced TRSs the left-hand side alone uniquely determines the employed rewrite rule. Left-reduce...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.05631302669644356, 0.05136846750974655, -0.008111212402582169, 0.019671406596899033, 0.005535921547561884, -0.010545339435338974, 0.018938880413770676, 0.01456660870462656, 0.04730904847383499, 0.042883362621068954, -0.06128810718655586, 0.005326082929968834, 0.01906096749007702, 0.0097...
fef02cfb5c5f2f3ab38bcba7d815f5dfcdf4acfb
subsection
45
59
Canonicity and Normalization Equivalence
Let \sqsupset be a total precedence and define s > t if and only if s _\mathcal {E}t and either d_\mathcal {R}(s) > d_\mathcal {R}(t) or both d_\mathcal {R}(s) = d_\mathcal {R}(t) and s \sqsupset _{\mathsf {lpo}} t.In the formalization we actually use \sqsupset _{\mathsf {kbo}} with all weights set to 1, since in contr...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.013167215511202812, 0.033780086785554886, -0.03448193147778511, 0.026609066873788834, -0.008132243528962135, -0.03866248577833176, 0.045741960406303406, 0.0015352851478382945, 0.027448227629065514, 0.05157032236456871, -0.024900227785110474, 0.012663718312978745, 0.02654803730547428, 0....
1a294ee2d5e65831b9ba37967aea4f412df84232
subsection
46
59
Canonicity and Normalization Equivalence
Correctness for Infinite Runs Completion as presented in the preceding sections does not always succeed in producing a finite complete presentation. It may fail because an unorientable equation is encountered or it may run forever. In the latter case it is possible that in the limit a possibly infinite complete present...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04088892042636871, 0.02753899246454239, -0.004100334830582142, 0.05486438423395157, -0.03222290799021721, -0.004054563585668802, -0.01144279446452856, 0.014898518100380898, 0.01775921694934368, 0.024716436862945557, -0.026165856048464775, -0.0006827533943578601, 0.002284744754433632, 0....
a5a6730fd1bd5aa0a4f4ee0ed7ee22da8df29752
subsection
47
59
Canonicity and Normalization Equivalence
After two orient steps, we apply deduce to generate the two critical pairs: \mathsf {a}\mathsf {b}\mathsf {a} &\rightarrow \mathsf {a}\mathsf {b} & \mathsf {b}\mathsf {b} &\rightarrow \mathsf {b} & \mathsf {a}\mathsf {b}\mathsf {a}\mathsf {b} &\approx \mathsf {a}\mathsf {b}\mathsf {b}\mathsf {a} & \mathsf {b}\mathsf {...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.044394418597221375, 0.02508055791258812, -0.016705116257071495, 0.026636650785803795, -0.006365488283336163, 0.030511626973748207, 0.017101766541600227, 0.01106046512722969, 0.03045060485601425, 0.013562418520450592, -0.028879255056381226, 0.013775999657809734, 0.017696743831038475, 0.0...
4017d6da40f0f671e6b9cd134b6b6ce35843b6b1
subsection
48
59
Canonicity and Normalization Equivalence
Since none of the rules \mathsf {a}\mathsf {b}\mathsf {a} \rightarrow \mathsf {a}\mathsf {b}^n survives, in the limit we obtain the TRS consisting of the single rule \mathsf {b}\mathsf {b} \rightarrow \mathsf {b}. This TRS is complete but not equivalent to \mathcal {E} as witnessed by non-joinability of \mathsf {a}\mat...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.01496393047273159, 0.03748231753706932, -0.006226703990250826, -0.005032489541918039, -0.0015127991791814566, -0.04682237282395363, 0.020877772942185402, 0.020969342440366745, 0.020313097164034843, 0.047280218452215195, -0.035834070295095444, 0.01341488491743803, 0.01307913102209568, 0....
dbd8af90a4b9f86a27f59d3ca773f0d27b0fb627
subsection
49
59
Ground Completion
In this section we focus on the special case of ground equations, that is, equations where both sides are ground terms.[Ground Completion linkbluecheck-square-o] The inference system KB_\mathsf {g} consists of the inference rules of KB_\mathsf {f} except for deduce.Snyder  proved that sets of ground equations can alway...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.029732907190918922, 0.03682669997215271, -0.011441601440310478, 0.0060640485025942326, 0.034751955419778824, -0.0946449264883995, 0.0500379353761673, -0.01162466686218977, 0.013882475905120373, 0.04137283191084862, -0.017116636037826538, 0.014782548882067204, 0.011029703542590141, 0.051...
7b681406e99044119facbb91f4dd1822bea2708b
subsection
50
59
Ground Completion
We start by applying \textsf {\small orient} to the last two equations:} \mathsf {f}(\mathsf {f}(\mathsf {f}(\mathsf {a}))) &\approx \mathsf {f}(\mathsf {b}) & \mathsf {f}(\mathsf {f}(\mathsf {b})) &\approx \mathsf {c} & \mathsf {f}(\mathsf {c}) &\mathrel {\leftarrow }\mathsf {a} & \mathsf {f}(\mathsf {a}) &\rightarrow...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.030484797433018684, 0.05965743586421013, -0.03835774585604668, -0.001669759745709598, -0.017027543857693672, -0.006770584732294083, 0.01864485628902912, 0.01830918714404106, 0.027265431359410286, 0.043819986283779144, -0.03225468471646309, 0.005492756143212318, 0.039303719997406006, 0.0...
f383d56d5c07489c4d7cf7df7a86bf2dca91ac09
subsection
51
59
Ground Completion
If s \mathrel {\leftarrow }\cdot \rightarrow t and the two contracted redexes are at parallel positions then trivially s \rightarrow \cdot \mathrel {\leftarrow }t. If the steps are identical then s = t. In the remaining case one of the contracted redexes is a subterm of the other contracted redex, and the effect of ded...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.0579441636800766, 0.0402161180973053, -0.01788061112165451, 0.0021702363155782223, -0.028209257870912552, 0.012060258537530899, 0.026973480358719826, 0.01022185105830431, 0.026775145903229713, 0.03554762899875641, -0.018597666174173355, 0.011945834383368492, 0.007105711847543716, 0.0503...
e704df69e96265a96f779ea4f41f6bfbdfd3ea89
subsection
52
59
Ground Completion
Now it is straightforward to verify that any infinite \vdash _\mathsf {g}-sequence would give rise to an infinite sequence P(\mathcal {E}_0,\varnothing ) \succ P(\mathcal {E}_1,\mathcal {R}_1) \succ \cdots , contradicting the well-foundedness of \succ .If > is total on \mathcal {E}-equivalent ground terms then every ma...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.0440068393945694, 0.05590882524847984, -0.007137378212064505, 0.007347188889980316, 0.002582578919827938, -0.007743921596556902, 0.026535330340266228, 0.00487142289057374, 0.016784854233264923, 0.05194149538874626, -0.03701823577284813, 0.008155914023518562, 0.03067050874233246, 0.02856...
0c4fe962ae1dd4206503467d0102d065ea00f8a1
subsection
53
59
Ground Completion
We formalized a new, short and direct proof of the following result due to van Oostrom . Here an element a is said to be complete if it is both terminating (there are no infinite rewrite sequences starting at a) and confluent (if b \mathrel {{\vphantom{\rightarrow }^{*}}{\mathrel {\leftarrow }}} a \rightarrow ^* c then...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.01786678656935692, 0.03967006504535675, -0.01687503606081009, 0.000027714213501894847, 0.010306588374078274, 0.0190263744443655, 0.029508426785469055, -0.01614266447722912, 0.029569456353783607, 0.041867177933454514, -0.027448633685708046, 0.021650701761245728, 0.04238593950867653, 0.03...
d7d10ce803506486ba4fecbd324bf9b2ad2caa30
subsection
54
59
Ground Completion
The remainder of the proof proceeds by induction on k together with Lemma .Right-reduced ground TRSs are terminating. linkbluecheck-square-o Let \mathcal {R} be a right-reduced ground TRS. For the sake of a contradiction, assume that \mathcal {R} is non-terminating. Then there is a minimal non-terminating term t (tha...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.02522602491080761, 0.03488608077168465, -0.010529919527471066, 0.01516155805438757, -0.007779168896377087, -0.015123406425118446, 0.033024270087480545, 0.009805033914744854, 0.022906390950083733, 0.057136259973049164, -0.044530875980854034, 0.026950489729642868, 0.02054097317159176, 0.0...
95c789e539df87be8795e1832d2e7a8cc8e9b1c9
subsection
55
59
Ground Completion
Both are basically handled by the following observation: d_\mathcal {R}(C[t\sigma ]) = d_\mathcal {R}(C[t{\downarrow }\sigma ]) + d_\mathcal {R}(t) for any term t (which holds due to random descent together with termination). This allows us to lift d_\mathcal {R}(s) = d_\mathcal {R}(t) and d_\mathcal {R}(s) > d_\mathca...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.021959489211440086, 0.02009773999452591, -0.0066000549122691154, -0.008568625897169113, 0.00991915725171566, -0.035495325922966, 0.0405922457575798, -0.0016709965420886874, 0.031069854274392128, 0.05621873214840889, -0.02519465982913971, 0.03105459362268448, 0.009446090087294579, 0.0352...
cd93136f1141b27bacb9a3e4c21f231c30cb964e
subsection
56
59
Correctness for Infinite Runs
Completion as presented in the preceding sections does not always succeed in producing a finite complete presentation. It may fail because an unorientable equation is encountered or it may run forever. In the latter case it is possible that in the limit a possibly infinite complete presentation is obtained. In this cas...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.03229685500264168, 0.02561158873140812, -0.004384343978017569, 0.0371505431830883, -0.02313895709812641, 0.0055557917803525925, -0.011424478143453598, 0.0244973786175251, 0.021521061658859253, 0.02713790535926819, -0.02281842939555645, 0.00720421364530921, -0.0009300990495830774, 0.0420...
c0e1e0f539ed841e575f9cf8d6e1daa966dca9c7
subsection
57
59
Correctness for Infinite Runs
After two orient steps, we apply deduce to generate the two critical pairs:\mathsf {a}\mathsf {b}\mathsf {a} &\rightarrow \mathsf {a}\mathsf {b} & \mathsf {b}\mathsf {b} &\rightarrow \mathsf {b} & \mathsf {a}\mathsf {b}\mathsf {a}\mathsf {b} &\approx \mathsf {a}\mathsf {b}\mathsf {b}\mathsf {a} & \mathsf {b}\mathsf {b}...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.04450254514813423, 0.025028865784406662, -0.016421377658843994, 0.02313643880188465, -0.010293884202837944, 0.03116399049758911, 0.01517756562680006, 0.012995170429348946, 0.030065162107348442, 0.013102000579237938, -0.029988855123519897, 0.011713203974068165, 0.019962046295404434, 0.01...
546fd691a333782b55b785768c9c6e6f6fd40eab
subsection
58
59
Correctness for Infinite Runs
Since none of the rules \mathsf {a}\mathsf {b}\mathsf {a} \rightarrow \mathsf {a}\mathsf {b}^n survives, in the limit we obtain the TRS consisting of the single rule \mathsf {b}\mathsf {b} \rightarrow \mathsf {b}. This TRS is complete but not equivalent to \mathcal {E} as witnessed by non-joinability of \mathsf {a}\mat...
{ "cite_spans": [] }
10.23638/LMCS-15(3:19)2019
1802.08437
Abstract Completion, Formalized
[ "Nao Hirokawa", "Aart Middeldorp", "Christian Sternagel", "Sarah Winkler" ]
[ "cs.LO" ]
2,018
en
Computer Science
[ -0.0155452536419034, 0.03758930787444115, -0.007635327987372875, -0.004397369455546141, -0.00137775344774127, -0.04527803137898445, 0.020884644240140915, 0.021067708730697632, 0.018901441246271133, 0.046712037175893784, -0.034294143319129944, 0.015499487519264221, 0.012631472200155258, 0.0...
36eebc33303501ffbe508b72476b2dd1fed91333
abstract
0
18
Abstract
Due to the fact that it is prohibitively expensive to completely annotate visual relationships, i.e., the (obj1, rel, obj2) triplets, relationship models are inevitably biased to object classes of limited pairwise patterns, leading to poor generalization to rare or unseen object combinations. Therefore, we are interest...
{ "cite_spans": [] }
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.09718739986419678, -0.013979845680296421, -0.03247719630599022, 0.021870216354727745, 0.016498049721121788, -0.014506379142403603, 0.04169534891843796, -0.005574387032538652, -0.00594067107886076, 0.012812314555048943, -0.03650632128119469, -0.03488856554031372, 0.0014670443488284945, 0...
9413a26fab0344a8adf2fab7820594b6f4b1f14a
subsection
1
18
Introduction
Thanks to the maturity of mid-level vision solutions such as object classification and detection , , , we are more ambitious to pursue higher-level vision-language tasks such as image captioning , , , , visual Q&A , , , and visual chatbot . Unfortunately, we gradually realize that many of the state-of-the-art systems m...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/cvpr.2016.90", "end": 240, "openalex_id": "https://openalex.org/W2194775991", "raw": "He, K., Zhang, X., Ren, S., Sun, J.: Deep residual learning for image recognition. In: CVPR (2016)", "source_ref_id": "77facf87265072cd14953...
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.08379649370908737, -0.01568514108657837, -0.042508564889431, 0.032560400664806366, 0.01299974787980318, 0.019270751625299454, 0.036832619458436966, -0.024488961324095726, 0.019224978983402252, 0.0024698758497834206, -0.02512979321181774, -0.05398251861333847, 0.004012832883745432, 0.030...
c6485ddba532c8355f12e16dfe3c64a873a515ed
subsection
2
18
Introduction
REF (b), “shuffle” is to discard the original one-to-one paired object alignments, and thus no explicit obj1-obj2 class information is used; “assemble” is to pose the relationship modeling into an unsupervised pair recover problem by transferring Region-of-Interest (ROI) features between the two unpaired domains. Our i...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/iccv.2017.244", "end": 1236, "openalex_id": "https://openalex.org/W2962793481", "raw": "Zhu, J.Y., Park, T., Isola, P., Efros, A.A.: Unpaired image-to-image translation using cycle-consistent adversarial networks. In: ICCV (2017)", ...
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.06822612881660461, -0.004607399459928274, -0.045463744550943375, 0.02009253390133381, -0.005477007944136858, 0.016888713464140892, 0.053152915090322495, 0.0016476793680340052, 0.02527967281639576, 0.008779995143413544, -0.016080129891633987, -0.002570684999227524, 0.002133973641321063, ...
ece0caa0be36b698ef846c1a30eb4280dfc2600d
subsection
3
18
Visual Relationships.
Modeling the object interactions such as verbs , , actions , , , and visual phrases , , , has been extensively studied in literature. In particular, our relationship model used in this paper follows the recent progress on modeling generic visual relationships, i.e., the (obj1, rel, obj2) triplets detected in images , ....
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1007/978-3-540-88682-2_3", "end": 134, "openalex_id": "https://openalex.org/W1584193343", "raw": "Gupta, A., Davis, L.S.: Beyond nouns: Exploiting prepositions and comparative adjectives for learning visual classifiers. In: ECCV (2008)",...
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.09744665026664734, -0.0032511453609913588, -0.0686458870768547, 0.02878550812602043, 0.004717497620731592, 0.04536730423569679, 0.04170618951320648, 0.018992027267813683, -0.014941920526325703, 0.005331497173756361, -0.04024174436926842, -0.03716030716896057, 0.01313424576073885, 0.0346...
262d4eb55173767df61a7f4c74871a4aa189cf4c
subsection
4
18
Method
Fig. REF illustrates the overview of using Shuffle-Then-Assemble to enhance the relationship model. The goal of the feature learning process is to pre-train the Object-Agnostic (OA) conv-layers, which result in the desired OA feature map for better relationship modeling. We will first introduce the widely-used relation...
{ "cite_spans": [] }
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.06993185728788376, -0.010145001113414764, -0.02851279266178608, -0.0008929317118600011, 0.0025229016318917274, -0.020244235172867775, 0.015377686358988285, 0.004508041776716709, 0.0001860473130363971, 0.001973698614165187, -0.010717087425291538, -0.0005377613706514239, -0.0163082797080278...
2db4cdae9124956ccc5c8bfd0612be08c5822e4f
subsection
5
18
Visual Relationship Model
The input of the visual relationship model is an image with a pair of object bounding boxes, and the output is an “obj1-rel-obj2” triplet, where “obj1” and “obj2” are the object classes of the two bounding boxes, and “rel” is the relationship class. In this paper, we adopt the common practice as in , that we do not dir...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1007/978-3-319-46448-0_51", "end": 564, "openalex_id": "https://openalex.org/W2479423890", "raw": "Lu, C., Krishna, R., Bernstein, M., Fei-Fei, L.: Visual relationship detection with language priors. In: ECCV (2016)", "source_ref_i...
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.09003668278455734, -0.03128392994403839, -0.061163902282714844, 0.04062332957983017, 0.006764196325093508, 0.03900572285056114, 0.015130740590393543, 0.03610623627901077, -0.0038895083125680685, -0.008202494122087955, -0.04639178141951561, -0.029574761167168617, 0.04156947880983353, 0.0...
02430d8383982a6bae5aa871a61614d140badc1d
subsection
6
18
Visual Relationship Model
For example, if most of the triplets of “stand on” is “person stand on street”, then the “stand on” classifier will mistakenly consider the joint pattern “person” and “street” into “stand on”, and fails in cases of “person stand on chair” or “dog stand on street”. [Figure: (a) The overview of unsupervised domain transf...
{ "cite_spans": [] }
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.07896015048027039, -0.010114099830389023, -0.033378053456544876, 0.024179257452487946, 0.01537709217518568, 0.02233339659869671, -0.0054880050010979176, 0.017924686893820763, -0.020990950986742973, 0.03173050656914711, -0.02440808340907097, -0.0027706988621503115, 0.003872876288369298, ...
c08e9c88834203f30fd3f0dffe7a48bb57d891ff
subsection
7
18
Shuffle-Then-Assemble Feature Learning
To alleviate the bias, we detail our proposed Shuffle-then-Assemble strategy to pre-train the Object-Agnostic (OA) conv-layers for obtaining the OA feature map. As discussed above, the bias is mainly due to the dominant object pairs in training data, therefore, our key idea is to discard the original one-to-one pairwis...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/iccv.2017.244", "end": 950, "openalex_id": "https://openalex.org/W2962793481", "raw": "Zhu, J.Y., Park, T., Isola, P., Efros, A.A.: Unpaired image-to-image translation using cycle-consistent adversarial networks. In: ICCV (2017)", ...
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.07806472480297089, 0.005889949854463339, -0.07757643610239029, 0.015838777646422386, -0.006557528395205736, 0.0010433298302814364, 0.027908597141504288, 0.010894881561398506, 0.018066581338644028, 0.023956531658768654, -0.018112357705831528, -0.010360818356275558, 0.009597871452569962, ...
1805d286f82c3bf82cb662afd14f0415477a8b15
subsection
8
18
Shuffle-Then-Assemble Feature Learning
Recall that there is one-to-one supervision between the two domains, we adopt the adversarial objective \mathcal {L}_{adv} such that the mapped features \lbrace F(\mathbf {a})\rbrace and \lbrace G(\mathbf {b})\rbrace are indistinguishable from B and A, respectively; in particular, the indistinguishability is measured b...
{ "cite_spans": [] }
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.0520440973341465, 0.010920102708041668, -0.03308844566345215, -0.01062249019742012, -0.004998370073735714, 0.0211534071713686, 0.03101278841495514, -0.005128098651766777, 0.027029354125261307, 0.03232533484697342, -0.04358883947134018, -0.014025960117578506, 0.00022034799621906132, 0.02...
93e97583bf84516055bbd36831f75899ce75a318
subsection
9
18
Shuffle-Then-Assemble Feature Learning
(REF ) together, the full objective for pre-training the OA conv-layers is:\phi ^* = \arg \min \limits _\phi \min \limits _{F, G}\max \limits _{D_A, D_B}\mathcal {L}_{adv}(A, B; \phi , F, G, D_A, D_B)+\lambda \mathcal {L}_{cycle}(A, B; \phi , F, G),where \lambda >0 is a trade-off hyper-parameter. Then, we can use \phi ...
{ "cite_spans": [] }
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.049934834241867065, -0.019671760499477386, -0.054848961532115936, -0.026585111394524574, 0.010827862657606602, 0.023731255903840065, 0.014826311729848385, 0.009172015823423862, -0.006352498661726713, 0.012407402507960796, -0.053811196237802505, -0.0017827056581154466, 0.008721808902919292...
cffe61f45c1cafe851d5d14dcffb97ca7a132419
subsection
10
18
Implementation Details
Network Architecture. For base CNN, we adopt Faster RCNN (VGG16) , which takes short width to be 600 and outputs the original 1/16\times 1/16 \times 512 feature map. As shown in Fig REF , our OA conv-layer has 1 filter of the size 1\times 1, stride 1, followed by a Leaky Relu . The transformation network is detailed in...
{ "cite_spans": [ { "arxiv_id": "", "doi": "10.1109/tpami.2016.2577031", "end": 165, "openalex_id": "https://openalex.org/W639708223", "raw": "Ren, S., He, K., Girshick, R., Sun, J.: Faster r-cnn: Towards real-time object detection with region proposal networks. In: NIPS (2015)", ...
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.06798789650201797, -0.032285094261169434, -0.04464375600218773, 0.0066904607228934765, 0.01187804713845253, 0.035214558243751526, 0.0024774540215730667, 0.02439691312611103, 0.010016619227826595, 0.03381085768342018, -0.05697190389037132, -0.010893931612372398, -0.005687273107469082, 0....
119312de61b1e113345132d8c50d16ea2a9df53d
subsection
11
18
Experiments
We evaluated our Shuffle-Then-Assemble method by performing visual relationship prediction on two benchmark datasets. We conducted experiments under extensive settings: supervised, weakly-supervised, and zero-shot, each of which has various ablative baselines and state-of-the-art methods. We also visualized qualitative...
{ "cite_spans": [] }
1808.00171
Shuffle-Then-Assemble: Learning Object-Agnostic Visual Relationship Features
[ "Xu Yang", "Hanwang Zhang", "Jianfei Cai" ]
[ "cs.CV" ]
2,018
en
Computer Science
[ -0.055528003722429276, 0.006803630385547876, -0.029015595093369484, 0.039989929646253586, -0.0007469493430107832, -0.015965444967150688, 0.052109017968177795, 0.002493646927177906, -0.0021025240421295166, 0.031686678528785706, -0.0018468631897121668, -0.0032835244201123714, 0.002304763300344...