pith. sign in

arxiv: 2606.28541 · v1 · pith:FVIQICQQnew · submitted 2026-06-26 · 🧮 math.OC · cs.CE· cs.NE

Comparing Scalar Objective Functions for Multi-Criteria Engineering Optimization

Olaf Frommann This is my paper

Pith reviewed 2026-06-30 01:03 UTC · model grok-4.3

classification 🧮 math.OC cs.CEcs.NE
keywords multi-criteria optimizationscalarizationPareto frontweighted sumsachievement functionsdesirability functionsfuzzy logicengineering design
0
0 comments X

The pith

Different scalar objective functions reach different parts of the Pareto front, with weighted sums structurally unable to access non-supported points on concave fronts.

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

This paper compares four scalar objective functions for turning bi-criteria engineering minimization into a single objective: weighted sums, achievement scalarizing functions, desirability functions, and a fuzzy-logic formulation. It evaluates them on two exact, analytically defined Pareto fronts—one convex and one concave—to separate the effect of the formulation itself from any numerical solver. Weighted sums prove simple but cannot reach interior non-supported regions when the front is concave. Achievement, desirability, and fuzzy methods can reach those regions, each through distinct mechanisms. Desirability applies nonlinear mappings to individual criteria, while fuzzy rules capture nonseparable preferences that depend on reference levels.

Core claim

Using analytically defined convex and concave Pareto fronts isolates the scalarization effect and shows that weighted sums are limited to supported points on concave fronts, while achievement scalarizing functions, desirability functions, and fuzzy formulations reach interior non-supported regions; desirability functions do so via nonlinear single-criterion preference mappings and fuzzy rules do so by expressing nonseparable, reference-dependent preferences.

What carries the argument

Analytically defined convex and concave Pareto fronts that isolate the scalar objective formulation from optimizer behavior.

If this is right

  • Weighted sums cannot reach non-supported regions on concave fronts.
  • Achievement scalarizing functions reach interior regions via reference-point mechanisms.
  • Desirability functions reach non-supported areas through nonlinear single-criterion mappings.
  • Fuzzy formulations reach them by expressing nonseparable and reference-dependent preferences.
  • Each formulation produces different densities of selected points depending on its parameters.

Where Pith is reading between the lines

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

  • Engineers working with concave trade-off surfaces should test scalarizations beyond weighted sums to avoid missing preferred designs.
  • The interpretability differences suggest fuzzy rules when engineering preferences involve interactions between criteria.
  • Real optimization runs with unknown front curvature may benefit from running multiple scalarizations in parallel.
  • The comparison framework could be extended to fronts with discontinuities or higher numbers of criteria.

Load-bearing premise

The two analytically defined Pareto fronts capture the structural differences that determine which scalarizations can reach which points in actual engineering problems.

What would settle it

If a weighted-sum formulation applied to the analytic concave front selects a non-supported interior point, the claimed structural limitation would be falsified.

Figures

Figures reproduced from arXiv: 2606.28541 by Olaf Frommann.

Figure 1
Figure 1. Figure 1: Analytical Pareto fronts used in this study. The convex front is defined by Equation [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Selected Pareto-front points on the convex front obtained by sweeping the native [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Selected Pareto-front points on the concave front obtained by sweeping the native [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: Parameter-induced selection density for the convex and concave Pareto fronts. The [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Sensitivity of the selected compromise to the native preference parameter of each [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Effect of the fuzzy rule base on the selected Pareto-front points for the convex test [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Effect of the fuzzy rule base on the selected Pareto-front points for the concave test [PITH_FULL_IMAGE:figures/full_fig_p012_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Parameter-induced selection density for the fuzzy rule sets F1–F4 on the convex and [PITH_FULL_IMAGE:figures/full_fig_p013_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: Sensitivity of the selected optima for the fuzzy rule sets F1–F4 to the reference [PITH_FULL_IMAGE:figures/full_fig_p014_9.png] view at source ↗
read the original abstract

Scalar objective functions are required when a multi-criteria optimization problem must yield a single preferred design rather than only a Pareto set. The choice of scalarization influences which compromise is selected, how preference parameters are interpreted, and whether non-supported Pareto regions can be reached. This paper compares four formulations for normalized bi-criteria minimization: weighted sums, achievement scalarizing functions, desirability functions, and a fuzzy-logic-based formulation. Two analytically defined Pareto fronts, one convex and one concave, isolate the effect of the objective formulation from numerical optimizer behavior. The comparison focuses on reachable Pareto regions, parameter-induced selection density, compensation between criteria, sensitivity, and interpretability. Results show that weighted sums are simple but structurally limited on concave fronts, while achievement, desirability, and fuzzy formulations reach interior non-supported regions through different mechanisms. Desirability functions introduce nonlinear single-criterion preference mappings, whereas fuzzy rules express nonseparable and reference-dependent engineering preferences.

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.

Referee Report

0 major / 3 minor

Summary. The manuscript compares four scalarizations for normalized bi-criteria minimization—weighted sums, achievement scalarizing functions, desirability functions, and a fuzzy-logic formulation—on two analytically defined Pareto fronts (one convex, one concave). The comparison isolates formulation effects from optimizer behavior and evaluates reachable Pareto regions, parameter-induced selection density, compensation, sensitivity, and interpretability. Results indicate weighted sums are limited on concave fronts while the other three reach interior non-supported points via distinct nonlinear or rule-based mechanisms.

Significance. If the analytical isolation and reported reachable sets hold, the work supplies a concrete, reproducible demonstration of how each scalarization encodes preferences and which regions of the front remain inaccessible under each choice. The use of exact, analytically defined fronts (rather than numerical optimization) is a methodological strength that directly supports the central claim about formulation effects.

minor comments (3)
  1. The abstract states that the fronts are 'analytically defined' but does not give their explicit functional forms; adding the equations (or a short appendix) would allow readers to reproduce the level-set evaluations without ambiguity.
  2. Parameter ranges and the exact normalization procedure applied to the criteria before scalarization are not stated in the abstract; these details are needed to interpret the reported selection densities.
  3. The fuzzy-logic formulation is described only at the level of 'rules'; a compact listing of the membership functions and inference rules used would improve clarity and reproducibility.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the constructive and positive assessment of the manuscript, including recognition of the methodological value of analytically defined Pareto fronts for isolating scalarization effects. No major comments were raised in the report.

Circularity Check

0 steps flagged

No significant circularity

full rationale

The paper is a comparative evaluation of four scalarization formulations on two analytically defined Pareto fronts (convex and concave). No derivation chain, fitted parameters, or predictions are present that reduce to the paper's own inputs by construction. The isolation of formulation effects follows directly from minimizing each scalar function exactly over the known front curves, and the weighted-sum limitation on concave fronts is the standard supporting-hyperplane property. No self-citations are load-bearing for the central claims.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are identifiable from the abstract alone; the work relies on standard concepts in multi-objective optimization.

pith-pipeline@v0.9.1-grok · 5684 in / 956 out tokens · 44038 ms · 2026-06-30T01:03:41.107008+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

15 extracted references · 8 canonical work pages

  1. [1]

    Decision-Making in a Fuzzy Environment

    Richard E. Bellman and Lotfi A. Zadeh. “Decision-Making in a Fuzzy Environment”. In: Management Science17.4 (1970), B141–B164.doi: 10.1287/mnsc.17.4.B141

    work page doi:10.1287/mnsc.17.4.b141 1970

  2. [2]

    A Closer Look at Drawbacks of Minimizing Weighted Sums of Objectives for Pareto Set Generation in Multicriteria Optimization Problems

    Indraneel Das and John E. Dennis. “A Closer Look at Drawbacks of Minimizing Weighted Sums of Objectives for Pareto Set Generation in Multicriteria Optimization Problems”. In: Structural and Multidisciplinary Optimization14 (1997), pp. 63–69.doi: 10.1007/ BF01197559

    1997

  3. [3]

    Simultaneous Optimization of Several Response Variables

    George Derringer and Ronald Suich. “Simultaneous Optimization of Several Response Variables”. In:Journal of Quality Technology12.4 (1980), pp. 214–219.doi: 10.1080/ 00224065.1980.11980968

    arXiv 1980

  4. [4]

    Multicriteria Optimization

    Matthias Ehrgott. Multicriteria Optimization. 2nd ed. Springer, 2005

    2005

  5. [5]

    Version 1.1.0

    Olaf Frommann.FuzzyGoal: A C++ Library for Fuzzy-Logic-Based Objective Functions. Version 1.1.0. 2026.doi: 10.5281/zenodo.20593012. url: https://github.com/of33/ FuzzyGoal. 16

    work page doi:10.5281/zenodo.20593012 2026

  6. [6]

    Technical Report

    Olaf Frommann.Objective Functions in Multi-Criteria Optimization: Weighting, Fuzzy Logic, and Solution-Space Topography. Technical Report. Version 1.0. Hochschule Bremen,

  7. [7]

    5281 / zenodo

    doi: 10 . 5281 / zenodo . 20585380. url: https : / / doi . org / 10 . 5281 / zenodo . 20585380

  8. [8]

    Comparing Scalar Objective Functions for Multi-Criteria Engineering Optimization

    Olaf Frommann.Reproducibility package for “Comparing Scalar Objective Functions for Multi-Criteria Engineering Optimization”. Version 1.0.0. 2026. doi: 10 . 5281 / zenodo . 20737015. url: https://doi.org/10.5281/zenodo.20737015

    work page doi:10.5281/zenodo.20737015 2026

  9. [9]

    The Desirability Function

    E. C. Harrington. “The Desirability Function”. In:Industrial Quality Control21.10 (1965), pp. 494–498

    1965

  10. [10]

    Fuzzy Multi-Objective Programming: A Systematic Literature Review

    N. Karimi et al. “Fuzzy Multi-Objective Programming: A Systematic Literature Review”. In: Expert Systems with Applications192 (2022), p. 116663.doi: 10.1016/j.eswa.2022. 116663

    work page doi:10.1016/j.eswa.2022 2022

  11. [11]

    The Weighted Sum Method for Multi-Objective Optimization: New Insights

    R. Timothy Marler and Jasbir S. Arora. “The Weighted Sum Method for Multi-Objective Optimization: New Insights”. In:Structural and Multidisciplinary Optimization41 (2010), pp. 853–862.doi: 10.1007/s00158-009-0460-7

    work page doi:10.1007/s00158-009-0460-7 2010

  12. [12]

    Springer Nature, 1998/2012.doi: 10.1007/978-1-4615-5563-6

    Kaisa Miettinen.Nonlinear Multiobjective Optimization. Springer Nature, 1998/2012.doi: 10.1007/978-1-4615-5563-6

    work page doi:10.1007/978-1-4615-5563-6 1998

  13. [13]

    On the Completeness and Constructiveness of Parametric Char- acterizations to Vector Optimization Problems

    Andrzej P. Wierzbicki. “On the Completeness and Constructiveness of Parametric Char- acterizations to Vector Optimization Problems”. In:Multiple Criteria Decision Making Theory and Application. Vol. 8. 1986, pp. 73–87.doi: 10.1007/BF01719738

    work page doi:10.1007/bf01719738 1986

  14. [14]

    Fuzzy Sets

    Lotfi A. Zadeh. “Fuzzy Sets”. In:Information and Control8.3 (1965), pp. 338–353.doi: 10.1016/S0019-9958(65)90241-X

    work page doi:10.1016/s0019-9958(65)90241-x 1965

  15. [15]

    Fuzzy Programming and Linear Programming with Several Objective Functions

    H.-J. Zimmermann. “Fuzzy Programming and Linear Programming with Several Objective Functions”. In:Fuzzy Sets and Systems 1.1 (1978), pp. 45–55.doi: 10 . 1016 / 0165 - 0114(78)90031-3. 17

    1978