Subdifferentials and Stability Analysis of Feasible Set and Pareto Front Mappings in Linear Multiobjective Optimization

Boris Mordukhovich; Juan Parra; Marco A. L\'opez; Mar\'ia J. C\'anovas

arxiv: 1907.09985 · v1 · pith:ICWD4INQnew · submitted 2019-07-23 · 🧮 math.OC

Subdifferentials and Stability Analysis of Feasible Set and Pareto Front Mappings in Linear Multiobjective Optimization

Mar\'ia J. C\'anovas , Marco A. L\'opez , Boris Mordukhovich , Juan Parra This is my paper

Pith reviewed 2026-05-24 17:09 UTC · model grok-4.3

classification 🧮 math.OC

keywords multiobjective linear optimizationPareto front mappingfeasible set mappingepigraphical multifunctionvector subdifferentialLipschitz stabilityright-hand side perturbationoptimal value computation

0 comments

The pith

Epigraphical mappings of feasible sets and Pareto fronts yield vector subdifferentials that certify Lipschitz stability under right-hand-side perturbations in linear multiobjective problems.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper studies linear multiobjective optimization problems whose inequality constraints are perturbed in their right-hand sides. It introduces epigraphical versions of the feasible-set and Pareto-front mappings by adding a fixed cone to their images so that directional changes become visible to subdifferential calculus. These epigraphical objects are then used to construct the corresponding vector subdifferentials and to obtain explicit Lipschitz moduli that quantify the stability of the perturbed mappings. In the ordinary linear-programming case the two subdifferentials are shown to be proportional sets, and the same construction supplies a direct method for computing optimal values without first locating an optimal solution.

Core claim

By defining epigraphical feasible-set and Pareto-front mappings through the addition of a fixed cone, the vector subdifferentials of these mappings can be described explicitly; the subdifferentials in turn certify Lipschitzian stability of the original mappings under right-hand-side perturbations and furnish the associated Lipschitz moduli. When the problem reduces to an ordinary linear program the subdifferentials of the two mappings become proportional, and the framework yields a procedure for recovering the optimal value without knowledge of any optimal solution.

What carries the argument

Epigraphical multifunction obtained by adding a fixed cone to the images of the original feasible-set or Pareto-front mapping; it converts directional variation into a form amenable to vector subdifferential calculus and stability estimates.

If this is right

The vector subdifferentials supply exact Lipschitz moduli for both the feasible-set and Pareto-front mappings under right-hand-side perturbations.
In ordinary linear programs the subdifferentials of the two mappings are proportional subsets.
Optimal values of linear programs can be recovered directly from the subdifferential construction without locating an optimal solution.
Stability statements hold in specific directions captured by the epigraphical augmentation.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The proportionality result for ordinary linear programs may simplify numerical computation of sensitivity information when moving from single- to multiobjective settings.
The same cone-augmentation device could be tested on problems whose constraint perturbations are measured in norms other than the Euclidean norm.
The method for computing optimal values without solutions may extend to detecting feasibility or boundedness in related linear systems.

Load-bearing premise

Adding a fixed cone to the images of the original mappings must preserve the structural properties required for the subdifferential calculus to apply.

What would settle it

For a concrete linear multiobjective instance, compute the subdifferential of the epigraphical feasible-set mapping at a reference point and check whether the resulting Lipschitz modulus equals the observed rate of change of the set under a sequence of right-hand-side perturbations; mismatch falsifies the claim.

read the original abstract

The paper concerns multiobjective linear optimization problems in R^n that are parameterized with respect to the right-hand side perturbations of inequality constraints. Our focus is on measuring the variation of the feasible set and the Pareto front mappings around a nominal element while paying attention to some specific directions. This idea is formalized by means of the so-called epigraphical multifunction, which is defined by adding a fixed cone to the images of the original mapping. Through the epigraphical feasible and Pareto front mappings we describe the corresponding vector subdifferentials, and employ them to verifying Lipschitzian stability of the perturbed mappings with computing the associated Lipschitz moduli. The particular case of ordinary linear programs is analyzed, where we show that the subdifferentials of both multifunctions are proportional subsets. We also provide a method for computing the optimal value of linear programs without knowing any optimal solution. Some illustrative examples are also given in the paper.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper introduces epigraphical multifunctions by adding a fixed cone to feasible-set and Pareto-front mappings, then uses the resulting vector subdifferentials to derive Lipschitz moduli under RHS perturbations in linear multiobjective problems.

read the letter

The punchline is that this paper supplies explicit Lipschitz moduli for the feasible-set and Pareto-front mappings in linear multiobjective optimization by defining epigraphical versions of those mappings and applying vector subdifferential calculus to them. The construction is applied to right-hand-side perturbations of the inequality constraints. For ordinary linear programs they further show that the subdifferentials of the two mappings are proportional subsets and give a method to recover the optimal value without an optimal solution. Examples are included to illustrate the claims. The epigraphical device and its direct application to the Pareto front look like the genuinely new technical step beyond the single-valued sensitivity results cited in the abstract. The linear setting lets them obtain concrete moduli rather than abstract existence statements, which is a clear plus for anyone who needs to compute stability constants. The soft spot is the one flagged in the stress test. Adding a fixed cone K to the images produces the epigraphical multifunction, but the paper must verify that the subdifferentials of the enlarged map recover the correct Aubin modulus for the original mappings. In the polyhedral linear case this probably holds, yet the text should state the precise qualification conditions (relative interior or recession-cone compatibility) under which the construction is valid. Without that check the central stability claim rests on an implicit assumption. The derivations appear to rest on standard variational-analysis results rather than circular fitting, which is fine. This is written for readers already comfortable with Mordukhovich coderivatives and set-valued analysis in parametric optimization. A specialist who wants explicit moduli for multiobjective problems will extract value; a general optimization reader will find it narrow. It deserves a serious referee because the technical step is checkable and the claims are specific enough to be tested against examples.

Referee Report

1 major / 1 minor

Summary. The paper studies right-hand-side parameterized linear multiobjective programs in R^n. It introduces epigraphical feasible-set and Pareto-front multifunctions obtained by adding a fixed cone to the images of the original mappings, derives the associated vector subdifferentials, and employs them to establish Lipschitzian stability of the perturbed mappings together with explicit Lipschitz moduli. A special case for ordinary linear programs is treated in which the subdifferentials of the two multifunctions are shown to be proportional; a method for computing optimal values without an optimal solution is also given, along with illustrative examples.

Significance. If the central derivations hold, the work supplies an explicit variational-analytic tool for sensitivity analysis of feasible-set and Pareto-front mappings in linear vector optimization. The computation of Lipschitz moduli via subdifferentials and the proportionality result for ordinary LPs are concrete contributions that could be used in perturbation studies.

major comments (1)

[Epigraphical multifunction definition] Epigraphical multifunction definition (abstract and § on epigraphical mappings): adding a fixed cone K to the images requires explicit qualification conditions (relative-interior or recession-cone compatibility) so that the subdifferential of the enlarged map yields the correct coderivative/Aubin modulus for the original unaugmented mappings. The manuscript does not state these conditions, which is load-bearing for the stability claims.

minor comments (1)

[Abstract] The abstract paragraph on epigraphical multifunctions is dense; a short clarifying sentence on why the cone addition preserves the directional variation properties needed for the stability results would improve readability.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading of the manuscript and the positive assessment of its contributions to variational analysis in linear multiobjective optimization. We address the single major comment below.

read point-by-point responses

Referee: Epigraphical multifunction definition (abstract and § on epigraphical mappings): adding a fixed cone K to the images requires explicit qualification conditions (relative-interior or recession-cone compatibility) so that the subdifferential of the enlarged map yields the correct coderivative/Aubin modulus for the original unaugmented mappings. The manuscript does not state these conditions, which is load-bearing for the stability claims.

Authors: We agree that qualification conditions are important for ensuring the subdifferential of the epigraphical multifunction correctly recovers the coderivative and Aubin modulus of the original mapping. In the linear setting of the paper, the polyhedral structure ensures that the recession cone of the epigraphical image coincides with that of the unaugmented mapping (both are generated by the same recession directions of the constraint system), so the relative-interior condition holds automatically when the nominal point is feasible. Nevertheless, to make the argument fully explicit and address the referee's concern, we will add a remark immediately after the definition of the epigraphical multifunctions stating the recession-cone compatibility condition and verifying that it is satisfied throughout the paper by linearity. This revision will not alter any proofs but will strengthen the presentation. revision: yes

Circularity Check

0 steps flagged

No circularity: standard application of subdifferential calculus to epigraphical mappings

full rationale

The paper constructs epigraphical feasible-set and Pareto-front mappings by adding a fixed cone K to the images of the original mappings, then invokes vector subdifferentials (from external variational analysis) to certify Lipschitz stability and compute moduli. No equation or step reduces by construction to a fitted parameter, self-definition, or load-bearing self-citation chain; the derivations remain self-contained against external benchmarks in convex and variational analysis. The ordinary LP case simply notes proportionality of subdifferentials without redefining inputs as outputs.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

The paper relies on standard results from variational analysis and convex analysis for subdifferential calculus; no free parameters or invented physical entities are introduced. The fixed cone added to form the epigraphical mapping is a modeling choice rather than a new entity with independent evidence.

axioms (2)

standard math Standard properties of subdifferentials and coderivatives in finite-dimensional spaces hold for the epigraphical multifunctions.
Invoked when the authors state that the subdifferentials describe the variation of the mappings.
domain assumption The problems are linear multiobjective programs with inequality constraints whose right-hand sides are the perturbation parameters.
Stated in the first sentence of the abstract.

pith-pipeline@v0.9.0 · 5707 in / 1472 out tokens · 18867 ms · 2026-05-24T17:09:53.587227+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

19 extracted references · 19 canonical work pages

[1]

T. Q. BAO and B. S. MORDUKHOVICH, Variational principles for set- valued mappings with applications to multiobjective optim ization, Control and Cybernetics 36 (2007), 531–562

work page 2007
[2]

T. Q. BAO and B. S. MORDUKHOVICH, Relative Pareto minimizers for multiobjective problems: existence and optimality condit ions, Math. Pro- gram. 122 (2010), 301–347

work page 2010
[3]

J. M. BOR WEIN and Q. J. ZHU, Techniques of Variational Analysis , Springer, New York, 2005

work page 2005
[4]

M. J. C ´ANOV AS, A. L. DONTCHEV, M. A. L ´OPEZ and J. PARRA, Metric regularity of semi-inﬁnite constraint systems , Math. Program. 104 (2005), 329–346

work page 2005
[5]

M. J. C ´ANOV AS, F. J. G´OMEZ-SENENT and J. PARRA, On the Lipschitz modulus of the argmin mapping in linear semi-inﬁnite optimi zation, Set- Valued Anal. 16 (2008), 511–538

work page 2008
[6]

M. J. C ´ANOV AS, D. KLATTE, M. A. L ´OPEZ and J. PARRA, Metric reg- ularity in convex semi-inﬁnite optimization under canonic al perturbations, SIAM J. Optim 18 (2007), 717–732

work page 2007
[7]

M. J. C ´ANOV AS, M. A. L´OPEZ, B. S. MORDUKHOVICH and J. PARRA, Variational analysis in semi-inﬁnite and inﬁnite programm ing, I: Stability of linear inequality systems of feasible solutions , SIAM J. Optim. 20 (2009), 1504–1526

work page 2009
[8]

M. J. C ´ANOV AS, M. A. L´OPEZ, B. S. MORDUKHOVICH and J. PARRA, Quantitative stability of linear inﬁnite inequality systems under block p er- turbations with applications to convex systems, TOP 20 (2012), 310–327

work page 2012
[9]

A. L. DONTCHEV and R. T. ROCKAFELLAR, Implicit Functions and Solution Mappings: A View from Variational Analysis , 2nd edition, Springer, New York, 2014

work page 2014
[10]

M. J. GISBERT, M. J. C ´ANOV AS, J. PARRA and F. J. TOLEDO, Lips- chitz modulus of the optimal value in linear programming , J. Optim. Theory Appl. 182 (2019), 133–152

work page 2019
[11]

M. A. GOBERNA and M. A. L ´OPEZ, Linear Semi-Inﬁnite Optimization , John Wiley & Sons, Chichester, UK, 1998

work page 1998
[12]

N. Q. HUY, B. S. MORDUKHOVICH and J. C. YAO, Coderivatives of frontier and solution maps in parametric multiobjective op timization, Tai- wanese J. Math. 12 (2008), 2083–2111

work page 2008
[13]

A. D. IOFFE, Variational Analysis of Regular Mappings , Springer, Cham, Switzerland, 2017. 22

work page 2017
[14]

KLATTE and B

D. KLATTE and B. KUMMER, Nonsmooth Equations in Optimization: Regularity, Calculus, Methods and Applications , Kluwer Academic, Dor- drecht, The Netherlands, 2002

work page 2002
[15]

B. S. MORDUKHOVICH, Complete characterizations of openness, met- ric regularity, and Lipschitzian properties of multifunct ions, Trans. Amer. Math. Soc. 340 (1993), 1–35

work page 1993
[16]

B. S. MORDUKHOVICH, Variational Analysis and Generalized Diﬀerenti- ation, I: Basic Theory, II: Applications , Springer, Berlin, 2006

work page 2006
[17]

B. S. MORDUKHOVICH, Variational Analysis and Applications , Springer, Cham, Switzerland, 2018

work page 2018
[18]

S. M. ROBINSON, Some continuity properties of polyhedral multifunc- tions, Mathematical Programming at Oberwolfach (Proc. Conf., Math. Forschungsinstitut, Oberwolfach, 1979). Math. Programming St ud. No. 14 (1981), 206–214

work page 1979
[19]

R. T. ROCKAFELLAR and R. J-B. WETS, Variational Analysis, Springer, Berlin, 1998. 23

work page 1998

[1] [1]

T. Q. BAO and B. S. MORDUKHOVICH, Variational principles for set- valued mappings with applications to multiobjective optim ization, Control and Cybernetics 36 (2007), 531–562

work page 2007

[2] [2]

T. Q. BAO and B. S. MORDUKHOVICH, Relative Pareto minimizers for multiobjective problems: existence and optimality condit ions, Math. Pro- gram. 122 (2010), 301–347

work page 2010

[3] [3]

J. M. BOR WEIN and Q. J. ZHU, Techniques of Variational Analysis , Springer, New York, 2005

work page 2005

[4] [4]

M. J. C ´ANOV AS, A. L. DONTCHEV, M. A. L ´OPEZ and J. PARRA, Metric regularity of semi-inﬁnite constraint systems , Math. Program. 104 (2005), 329–346

work page 2005

[5] [5]

M. J. C ´ANOV AS, F. J. G´OMEZ-SENENT and J. PARRA, On the Lipschitz modulus of the argmin mapping in linear semi-inﬁnite optimi zation, Set- Valued Anal. 16 (2008), 511–538

work page 2008

[6] [6]

M. J. C ´ANOV AS, D. KLATTE, M. A. L ´OPEZ and J. PARRA, Metric reg- ularity in convex semi-inﬁnite optimization under canonic al perturbations, SIAM J. Optim 18 (2007), 717–732

work page 2007

[7] [7]

M. J. C ´ANOV AS, M. A. L´OPEZ, B. S. MORDUKHOVICH and J. PARRA, Variational analysis in semi-inﬁnite and inﬁnite programm ing, I: Stability of linear inequality systems of feasible solutions , SIAM J. Optim. 20 (2009), 1504–1526

work page 2009

[8] [8]

M. J. C ´ANOV AS, M. A. L´OPEZ, B. S. MORDUKHOVICH and J. PARRA, Quantitative stability of linear inﬁnite inequality systems under block p er- turbations with applications to convex systems, TOP 20 (2012), 310–327

work page 2012

[9] [9]

A. L. DONTCHEV and R. T. ROCKAFELLAR, Implicit Functions and Solution Mappings: A View from Variational Analysis , 2nd edition, Springer, New York, 2014

work page 2014

[10] [10]

M. J. GISBERT, M. J. C ´ANOV AS, J. PARRA and F. J. TOLEDO, Lips- chitz modulus of the optimal value in linear programming , J. Optim. Theory Appl. 182 (2019), 133–152

work page 2019

[11] [11]

M. A. GOBERNA and M. A. L ´OPEZ, Linear Semi-Inﬁnite Optimization , John Wiley & Sons, Chichester, UK, 1998

work page 1998

[12] [12]

N. Q. HUY, B. S. MORDUKHOVICH and J. C. YAO, Coderivatives of frontier and solution maps in parametric multiobjective op timization, Tai- wanese J. Math. 12 (2008), 2083–2111

work page 2008

[13] [13]

A. D. IOFFE, Variational Analysis of Regular Mappings , Springer, Cham, Switzerland, 2017. 22

work page 2017

[14] [14]

KLATTE and B

D. KLATTE and B. KUMMER, Nonsmooth Equations in Optimization: Regularity, Calculus, Methods and Applications , Kluwer Academic, Dor- drecht, The Netherlands, 2002

work page 2002

[15] [15]

B. S. MORDUKHOVICH, Complete characterizations of openness, met- ric regularity, and Lipschitzian properties of multifunct ions, Trans. Amer. Math. Soc. 340 (1993), 1–35

work page 1993

[16] [16]

B. S. MORDUKHOVICH, Variational Analysis and Generalized Diﬀerenti- ation, I: Basic Theory, II: Applications , Springer, Berlin, 2006

work page 2006

[17] [17]

B. S. MORDUKHOVICH, Variational Analysis and Applications , Springer, Cham, Switzerland, 2018

work page 2018

[18] [18]

S. M. ROBINSON, Some continuity properties of polyhedral multifunc- tions, Mathematical Programming at Oberwolfach (Proc. Conf., Math. Forschungsinstitut, Oberwolfach, 1979). Math. Programming St ud. No. 14 (1981), 206–214

work page 1979

[19] [19]

R. T. ROCKAFELLAR and R. J-B. WETS, Variational Analysis, Springer, Berlin, 1998. 23

work page 1998