Prediction-Augmented Mechanism Design for Weighted Facility Location

Yangguang Shi; Zhenyu Xue

REVIEW 3 major objections 3 minor 32 references

Prediction-Augmented Mechanism Design for Weighted Facility Location

T0 review · 3 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper proves that a weight-ignoring median mechanism with phantom prediction copies achieves sharp tunable consistency and robustness bounds for weighted facility location, and proves an impossibility for perfect consistency.

desk verdict The weighted CMP bounds for COA instances are correct, but the reduction that lifts them to all weighted instances fails at Lemma 6, leaving Theorem 1 unsupported as written. read the letter →

arxiv 2507.06509 v3 pith:VZHJSRQZ submitted 2025-07-09 cs.DS cs.GTcs.LG

Yangguang Shi , Zhenyu Xue This is my paper · ORCID

classification cs.DScs.GTcs.LG MSC 91B1468W2590B80

keywords weightedfacilitylocationprediction-augmentedmechanismdesignstrategyproofconsistencyandrobustnesscoordinatemedianwithpredictionapproximationimpossibilityresultlearning-augmentedalgorithms

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper extends prediction-augmented facility location from unweighted to weighted agents, where agents matter unequally. It claims that a simple mechanism—the coordinate-wise median computed after inserting copies of the predicted optimal location—remains strategyproof and reaches a tunable balance between consistency (efficiency when predictions are right) and robustness (loss when predictions are wrong), with both bounds expressed through the smallest and largest agent weights. The worst-case analysis is carried out by reducing arbitrary weighted instances to a family of symmetric clusters-on-axes instances. The paper also proves that no deterministic strategyproof mechanism can be simultaneously 1-consistent and robust within a factor that grows gently with the number of agents and the weight ratio, so the trade-off is not an artifact of the proposed mechanism.

What carries the argument

The machinery is the family of weighted Clusters-and-OPT-on-Axes (COA) instances: agents placed at $(0,1)$ with maximum weight, agents split symmetrically at $(\pm x,0)$ with minimum weight, the prediction at the origin, and the optimum at $(0,1)$. The paper argues that every configuration where CMP outputs the origin can be transformed, by moving points onto the axes and rearranging them symmetrically, into a COA instance whose approximation ratio is at least as large; the consistency and robustness formulas then come from a one-variable calculus maximization over $x$. The transformation lemmas are inherited from the unweighted proof and are restated in the weighted setting; they are what carries the argument from arbitrary instances to the clean COA worst case.

What would settle it

Construct a small weighted instance with one high-weight agent placed off the axes that the paper's transformation rules cannot move onto the axes without lowering the approximation ratio, then compute CMP's ratio; if it exceeds $\frac{\sqrt{(1+c)^2 W_{\min}^2 + (1-c)^2 W_{\max}^2}}{(1+c) W_{\min}}$ for accurate predictions, the claimed worst-case reduction fails.

Watch

Extended reading notes

Core claim

On its own terms, the paper's central discovery is that the Coordinate Median with Prediction (CMP) mechanism—add $m = \lfloor cn\rfloor$ phantom copies of the predicted optimum $\hat{o}$ to the reported locations and take the coordinate-wise median—extends cleanly to weighted agents even though it never looks at the weights. The mechanism is strategyproof by construction, and the paper claims it is $\alpha(c)$-consistent and $\beta(c)$-robust with $\alpha(c) = \frac{\sqrt{(1+c)^2 W_{\min}^2 + (1-c)^2 W_{\max}^2}}{(1+c) W_{\min}}$ and $\beta(c) = \frac{\sqrt{(1-c)^2 W_{\min}^2 + (1+c)^2 W_{\max}^2}}{(1-c) W_{\min}}$ in $\mathbb{R}^2$. At $c=0$ the bound reduces to the known weighted coordinate-median approximation $W_{\max}/W_{\min}$, and when all weights coincide it reduces to the known unweighted bounds. A second result shows this trade-off is inherent: no deterministic strategyproof mechanism can reach 1-consistency while keeping robustness bounded by $O(n \cdot W_{\max}/W_{\min})$.

Load-bearing premise

The load-bearing premise is that every weighted instance can be transformed into a symmetric clusters-on-axes instance without lowering CMP's approximation ratio; if that reduction fails for weights, the claimed bounds only hold for the special family.

Editorial extensions

If this is right

Tuning $c\in[0,1)$ moves a facility planner along a consistency-robustness frontier: larger $c$ improves performance under accurate predictions, while smaller $c$ reins in loss under misleading predictions.
Because the mechanism ignores weights, it can be deployed in weighted settings without soliciting or verifying agent weights, preserving its strategyproofness; the cost is a trade-off bound that degrades as $W_{\max}/W_{\min}$ grows.
Setting $c=0$ reproduces the best-known deterministic weighted facility-location guarantee, so prediction-augmented operation is free in the worst case.
The impossibility result implies weighted prediction-augmented design must accept an inherent factor $\Omega(n W_{\max}/W_{\min})$ if it insists on perfect consistency.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

A complete proof of the weighted transformation lemmas, particularly the step that redistributes points on the $x$-axis by their weights, would turn the COA worst-case calculation into a full worst-case bound; readers should treat the COA-only guarantee as the currently established part.
Since CMP never uses weights, a natural extension the paper leaves open is to add weighted phantom points or use a weighted median, which might improve the constants while keeping strategyproofness.
The impossibility proof assumes anonymity and unanimity through the characterization it cites; dropping anonymity could in principle open different trade-offs, though the paper does not explore that.
The coordinate-wise structure limits the result to $\mathbb{R}^2$; in general metric spaces, where median mechanisms are not strategyproof, the same consistency-robustness balance would need a different mechanism.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

Quick take: this is worth reading for the problem and the COA analysis, but not for the universal guarantees as stated. Theorem 1 claims consistency and robustness bounds for all weighted instances in R^2, and the proof rests on reducing arbitrary instances to weighted COA instances. That reduction is where it falls apart.

What is genuinely new: the weighted extension of prediction-augmented facility location is natural, and the bounds depending on Wmax/Wmin are exactly what you would want. The COA computations in Theorems 4 and 13 are correct: maximizing over x gives the stated closed forms. The impossibility result in Theorem 2 is also a sensible target, though its proof is a sketch that leans on an unweighted characterization.

The load-bearing problem is Lemma 6. It says that if |A_-| >= |A_+ ∪ {o}| then you can shift non-y-axis points left by epsilon and the mechanism cost does not increase. In the unweighted case that is just a count. With weights, the cost difference is epsilon(Σ_{A_-} w_i − Σ_{A_+} w_i), and the cardinality condition says nothing about the weighted sum. One heavy point on the right flips the sign. Lemma 8 has the same disease in another form: it re-assigns average weights to x-axis points and invokes an unweighted corollary, but weights are part of the input and changing them changes the optimal cost. Lemma 10 is imported from Agrawal et al. without proof. So Theorem 5 has no support in the weighted case. The COA bounds are fine, but they are bounds for a special family, not for all instances.

There is also a smaller point: the text is rough in places, with duplicated and garbled sentences, and the impossibility section reads more like a construction than a proof. Both are fixable.

If a proper weighted reduction is supplied, the result is likely true and useful. As submitted, Theorem 1 is unsupported, but the paper is serious and the problem is worth refereeing. I would send it to review, and expect major revision.

Referee Report

3 major / 3 minor

Summary. The paper studies deterministic, strategyproof mechanisms for the weighted facility location problem in R^2 when the designer receives a prediction of the optimal location. It proposes to run the unweighted Coordinate Median with Prediction (CMP) mechanism, which adds cn phantom copies of the prediction and takes the coordinate-wise median, and claims a consistency-robustness trade-off (Theorem 1). The proof analyzes a family of worst-case instances (weighted COA) and claims via a reduction that every weighted instance can be transformed into a COA instance without lowering the approximation ratio. The paper also states an impossibility result (Theorem 2) for mechanisms that are 1-consistent and have bounded robustness. The COA-family computations are correct, but the reduction to COA is not validly proved for non-uniform weights, so the universal upper-bound claim is not established as written.

Significance. The target result is a natural and valuable extension of Agrawal et al.'s unweighted learning-augmented facility location to weighted agents; if established, it would provide an explicit tunable trade-off and would recover known unweighted bounds as a special case. The COA computations in Theorems 4 and 13 are elementary and check out, and CMP's strategyproofness is immediate because the mechanism ignores weights and uses only phantom points independent of reports. The missing piece is a valid weighted analogue of the reduction to COA; without it the main theorem is not proven for arbitrary weighted instances. The paper would be a useful contribution if that gap can be repaired, but the universal claims should not be cited as proven in the current form.

major comments (3)

[3.2, Lemma 6] The reduction step in Lemma 6 is invalid for non-uniform weights. The proof shifts all non-y-axis points left by epsilon and asserts C_u(f(P)) - C_u(f(Q)) = epsilon( sum_{A_-} w_i - sum_{A_+} w_i ) <= 0 from the cardinality condition |A_-| >= |A_+| + 1. Cardinality does not imply the weighted inequality; one heavy point in A_+ can make the sum positive, so the mechanism cost can decrease and r(Q) can fall below r(P). Since Lemma 6 is the first step in Theorem 5's claim that all instances reduce to COA, Theorem 1's bounds are not established for arbitrary weighted instances.
[3.2, Lemma 8] The OA-to-COA transformation reassigns weights by averaging them over the x-axis points and then invokes an unweighted corollary (Corollary 3 in [28]) to conclude that C_U(P, o(P)) decreases while C_U(P, (0,0)) remains unchanged. With non-uniform weights, this redistribution does not preserve the weighted cost to the optimal location, so the reduction is unsupported. A correct weighted analogue of the Agrawal et al. reduction is needed before Theorem 5 can be used to prove the claimed worst-case bounds.
[3.5, Theorem 2] The impossibility proof applies Corollary 3 of [24], a characterization of deterministic, strategyproof, anonymous, and unanimous mechanisms as GCM with n-1 constant points, to prediction-augmented mechanisms. The prediction o(P) is a function of the reported profile, not a constant phantom point, and the paper does not verify that the mechanisms under consideration satisfy the hypotheses of that characterization. Therefore the claimed n * Wmax/Wmin lower bound is not proven. The proof also does not specify the wrong prediction used in the robustness scenario.

minor comments (3)

[3.1] There are several typos in the text: 'heAs before' and 'TThe' appear in the paragraph after Algorithm 1, and the sentence 'The procedure of our mechanism does not rely on the weights, consider weights when computing the facility location' is garbled.
[3.2] Lemma 11's proof cites 'Lemma 5' where the statement being used is Lemma 10; the citations to 'Corollary 3 in [28]' (Lemma 8) and 'Corollary 7 in [28]' (Lemma 12) should be checked and harmonized.
[3.4] In the three-agent example, the consistency ratio for c < 2/3 is 3/sqrt(2), not 3*sqrt(2) as printed.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity found: the claimed bounds are derived for explicit COA instances and the reduction, while incomplete, is not a re-labeling of the paper's own inputs.

full rationale

The paper's main result is not circular in the sense of deriving X from X. CMP is an explicit mechanism with confidence parameter c; the consistency and robustness expressions in Theorems 4 and 13 are computed by optimizing over weighted COA instances, not by fitting constants or renaming known quantities. Strategyproofness is inherited from the coordinate median, an external result, and the weighted cost is defined independently of the mechanism. The load-bearing reduction (Theorem 5) is admittedly borrowed: 'Many lemmas used in the proof are the same as [1]. For those lemmas, we just provide them and do not give the detailed proofs.' This reliance on Agrawal et al. is external, not self-citational, and the cited work is not the present paper's own prior claim. There are serious rigor gaps: Lemma 6 replaces a weighted sum condition with a cardinality condition, so the inequality C_u(f(P))-C_u(f(Q)) = epsilon(sum_{A_-}w_i - sum_{A_+}w_i) <= 0 does not follow from |A_-| >= |A_+| + 1 with non-uniform weights; Lemma 8's weight redistribution is also asserted rather than proved. These are correctness concerns about the reduction from arbitrary instances to COA, not circularity: the COA bounds themselves are computed directly and would stand if the reduction were repaired. The impossibility proof invokes Peters et al.'s external characterization but does not restate the target theorem as its premise. Accordingly, the correct finding is 'no significant circularity' (score 0).

Assumptions & free parameters 1 free parameters · 3 assumptions · 0 invented entities

The central upper bound rests on an assumed extension of the unweighted COA reduction to weighted agents, and the lower bound rests on a characterization theorem applied in a new context. No entity is invented and no data fitting is performed.

free parameters (1)

c = c in [0,1), chosen by mechanism designer
Explicit trade-off parameter controlling the number of phantom prediction copies. Not fitted to data; part of the mechanism definition.

assumptions (3)

domain assumption The CA/OA reduction lemmas of Agrawal et al. [1] extend to weighted agents without changing their conclusions.
Used throughout Section 3.2 to transform arbitrary instances into COA. Not proven for weighted case; Lemma 8's redistribution step appears to fail for non-uniform weights.
domain assumption Any deterministic, strategyproof, anonymous, and unanimous mechanism can be expressed as a GCM mechanism with n-1 constant points, and this representation holds for prediction-augmented mechanisms with prediction-dependent phantoms.
Invoked in Section 3.5 to prove Theorem 2. Applicability to prediction-augmented mechanisms, where the function of the prediction may violate unanimity, is not justified.
domain assumption The weighted geometric median in the COA instances is exactly at (0,1) for the extremal x values used in Theorems 4 and 13.
Needed to compute the optimal cost in the denominator. Holds under the stated weight conditions, but the paper does not verify the subgradient condition explicitly.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Prediction-Augmented Mechanism Design for Weighted Facility Location." pith.science (2026). https://pith.science/paper/VZHJSRQZ

@misc{pith2026250706509,
  author       = {Pith},
  title        = {Pith review of: Prediction-Augmented Mechanism Design for Weighted Facility Location},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VZHJSRQZ}},
  note         = {Machine review of arXiv:2507.06509}
}

abstract

Facility location is fundamental in operations research, mechanism design, and algorithmic game theory, with applications ranging from urban infrastructure planning to distributed systems. Recent research in this area has focused on augmenting classic strategyproof mechanisms with predictions to achieve an improved performance guarantee against the uncertainty under the strategic environment. Previous work has been devoted to address the trade-off obstacle of balancing the consistency (near-optimality under accurate predictions) and robustness (bounded inefficiency under poor predictions) primarily in the unweighted setting, assuming that all agents have the same importance. However, this assumption may not be true in some practical scenarios, leading to research of weighted facility location problems. The major contribution of the current work is to provide a prediction augmented algorithmic framework for balancing the consistency and robustness over strategic agents with non-uniform weights. In particular, through a reduction technique that identifies a subset of representative instances and maps the other given locations to the representative ones, we prove that there exists a strategyproof mechanism achieving a bounded consistency guarantee of $\frac{\sqrt{(1+c)^2W^2_{\min}+(1-c)^2W^2_{\max}}}{(1+c)W_{\min}}$ and a bounded robustness guarantee of $\frac{\sqrt{(1-c)^2W^2_{\min}+(1+c)^2W^2_{\max}}}{(1-c)W_{\min}}$ in weighted settings, where $c$ can be viewed as a parameter to make a trade-off between the consistency and robustness and $W_{\min}$ and $W_{\max}$ denote the minimum and maximum agents' weight. We also prove that there is no strategyproof deterministic mechanism that reach $1$-consistency and $O\left( n \cdot \frac{W_{\max}}{W_{\min}} \right)$-robustness in weighted FLP, even with fully predictions of all agents.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

32 extracted references · 27 canonical work pages

[28]

Optimality of the coordinate-wise median mechanism for strategyproof facility location in two dimensions

Goel Sumit and Hann-Caruthers Wade. Coordinate-wise median: Not bad, not bad, pretty good.arXiv preprint arXiv:2007.00903, 2020

work page Pith review arXiv 2007
[24]

Range convexity, continuity, and strategy-proofness of voting schemes.Zeitschrift für Operations Research, 38:213–229, 1993

Hans Peters, Hans van der Stel, and Ton Storcken. Range convexity, continuity, and strategy-proofness of voting schemes.Zeitschrift für Operations Research, 38:213–229, 1993

work page 1993
[1]

Learning- augmented mechanism design: Leveraging predictions for facility location

Priyank Agrawal, Eric Balkanski, Vasilis Gkatzelis, Tingting Ou, and Xizhi Tan. Learning- augmented mechanism design: Leveraging predictions for facility location. InProceedings of the 23rd ACM Conference on Economics and Computation, pages 497–528, 2022

work page 2022
[2]

A survey of healthcare facility location.Computers & Operations Research, 79:223–263, 2017

Amir Ahmadi-Javid, Pardis Seyedi, and Siddhartha S Syam. A survey of healthcare facility location.Computers & Operations Research, 79:223–263, 2017. 15

work page 2017
[3]

Secretary and online matching problems with machine learned advice.Advances in Neural Information Processing Systems, 33:7933–7944, 2020

Antonios Antoniadis, Themis Gouleakis, Pieter Kleer, and Pavel Kolev. Secretary and online matching problems with machine learned advice.Advances in Neural Information Processing Systems, 33:7933–7944, 2020

2020
[4]

Randomized strategic facility location with predictions.Advances in Neural Information Processing Systems, 37:35639– 35664, 2024

Eric Balkanski, Vasilis Gkatzelis, and Golnoosh Shahkarami. Randomized strategic facility location with predictions.Advances in Neural Information Processing Systems, 37:35639– 35664, 2024

work page 2024
[5]

Facility location optimization model for emergency humanitarian logistics.International Journal of Disaster Risk Reduc- tion, 24:485–498, 2017

Chawis Boonmee, Mikiharu Arimura, and Takumi Asada. Facility location optimization model for emergency humanitarian logistics.International Journal of Disaster Risk Reduc- tion, 24:485–498, 2017

work page 2017
[6]

Cambridge University Press, 2016

Felix Brandt, Vincent Conitzer, Ulle Endriss, Jérôme Lang, and Ariel D Procaccia.Hand- book of computational social choice. Cambridge University Press, 2016

work page 2016

Show all 32 references

[7]

Temporary facility location problem in humanitarian logistics: A systematic literature review.Logistics, 9(1):42, 2025

María Fernanda Carnero Quispe, Lucciana Débora Chambilla Mamani, Hugo Tsug- unobu Yoshida Yoshizaki, and Irineu de Brito Junior. Temporary facility location problem in humanitarian logistics: A systematic literature review.Logistics, 9(1):42, 2025

2025
[8]

Mechanism design for facility location problems: a survey.arXiv preprint arXiv:2106.03457, 2021

Hau Chan, Aris Filos-Ratsikas, Bo Li, Minming Li, and Chenhao Wang. Mechanism design for facility location problems: a survey.arXiv preprint arXiv:2106.03457, 2021

2021 arXiv
[9]

Strategic facility location via predictions

Qingyun Chen, Nick Gravin, and Sungjin Im. Strategic facility location via predictions. arXiv preprint arXiv:2410.07497, 2024

2024 arXiv
[10]

Network and discrete location: models, algorithms and applications.Journal of the Operational Research Society, 48(7):763–764, 1997

Mark Daskin. Network and discrete location: models, algorithms and applications.Journal of the Operational Research Society, 48(7):763–764, 1997

1997
[11]

Incentive compatible regression learning

Ofer Dekel, Felix Fischer, and Ariel D Procaccia. Incentive compatible regression learning. Journal of Computer and System Sciences, 76(8):759–777, 2010

2010
[12]

Multiple obnoxious facilities location: A cooperative model.IISE Transactions, 52(12):1403–1412, 2020

Tammy Drezner, Zvi Drezner, and Pawel Kalczynski. Multiple obnoxious facilities location: A cooperative model.IISE Transactions, 52(12):1403–1412, 2020

2020
[13]

Strategy-proof mechanisms for facility location games with many facilities

Bruno Escoffier, Laurent Gourves, Nguyen Kim Thang, Fanny Pascual, and Olivier Span- jaard. Strategy-proof mechanisms for facility location games with many facilities. InAl- gorithmic Decision Theory: Second International Conference, ADT 2011, Piscataway, NJ, USA, October 26-2...

2011
[14]

Strategyproof facility location and the least squares ob- jective

Michal Feldman and Yoav Wilf. Strategyproof facility location and the least squares ob- jective. InProceedings of the fourteenth ACM conference on Electronic commerce, pages 873–890, 2013

2013
[15]

Learning augmented online facility location.arXiv preprint arXiv:2107.08277, 2021

Dimitris Fotakis, Evangelia Gergatsouli, Themis Gouleakis, and Nikolas Patris. Learning augmented online facility location.arXiv preprint arXiv:2107.08277, 2021

2021 arXiv
[16]

Median location problem.Facility location: Concepts, models, algo- rithms and case studies, pages 177–191, 2009

Masoomeh Jamshidi. Median location problem.Facility location: Concepts, models, algo- rithms and case studies, pages 177–191, 2009

2009
[17]

Online facility location with predictions

Shaofeng H-C Jiang, Erzhi Liu, You Lyu, Zhihao Gavin Tang, and Yubo Zhang. Online facility location with predictions. InInternational Conference on Learning Representations, 2022

2022
[18]

Online algorithms for weighted paging with predictions.ACM Transactions on Algorithms (TALG), 18(4):1–27, 2022

Zhihao Jiang, Debmalya Panigrahi, and Kevin Sun. Online algorithms for weighted paging with predictions.ACM Transactions on Algorithms (TALG), 18(4):1–27, 2022. 16

2022
[19]

Proportional fairness in obnoxious facility location

Alexander Lam, Haris Aziz, Bo Li, Fahimeh Ramezani, and Toby Walsh. Proportional fairness in obnoxious facility location. InProceedings of the 23rd International Conference on Autonomous Agents and Multiagent Systems, pages 1075–1083, 2024

2024
[20]

Competitive caching with machine learned advice

Thodoris Lykouris and Sergei Vassilvitskii. Competitive caching with machine learned advice. InProceedings of the 35th International Conference on Machine Learning (ICML), pages 3296–3305. PMLR, 2018

2018
[21]

Competitive caching with machine learned advice.Journal of the ACM (JACM), 68(4):1–25, 2021

Thodoris Lykouris and Sergei Vassilvitskii. Competitive caching with machine learned advice.Journal of the ACM (JACM), 68(4):1–25, 2021

2021
[22]

Strategyproof facility location for three agents on a circle

Reshef Meir. Strategyproof facility location for three agents on a circle. InInternational symposium on algorithmic game theory, pages 18–33. Springer, 2019

2019
[23]

A healthcare facility location selection problem with fuzzy topsis method for a regional hospital.Avrupa Bilim ve Teknoloji Dergisi, (16):750–757, 2019

Pınar Miç and Z Figen Antmen. A healthcare facility location selection problem with fuzzy topsis method for a regional hospital.Avrupa Bilim ve Teknoloji Dergisi, (16):750–757, 2019

2019
[25]

Approximate mechanism design without money

Ariel D Procaccia and Moshe Tennenholtz. Approximate mechanism design without money. ACM Transactions on Economics and Computation (TEAC), 1(4):1–26, 2013

2013
[26]

Improving online algorithms via ml predictions.Advances in Neural Information Processing Systems, 31, 2018

Manish Purohit, Zoya Svitkina, and Ravi Kumar. Improving online algorithms via ml predictions.Advances in Neural Information Processing Systems, 31, 2018

2018
[27]

Strategy-proof location on a network.Journal of Economic Theory, 104(2):405–428, 2002

James Schummer and Rakesh V Vohra. Strategy-proof location on a network.Journal of Economic Theory, 104(2):405–428, 2002

2002
[29]

False-name-proof mechanism design with- out money

Taiki Todo, Atsushi Iwasaki, and Makoto Yokoo. False-name-proof mechanism design with- out money. InThe 10th International Conference on Autonomous Agents and Multiagent Systems-Volume 2, pages 651–658, 2011

2011
[30]

Online algorithms for multi-shop ski rental with machine learned advice.Advances in Neural Information Processing Systems, 33:8150–8160, 2020

Shufan Wang, Jian Li, and Shiqiang Wang. Online algorithms for multi-shop ski rental with machine learned advice.Advances in Neural Information Processing Systems, 33:8150–8160, 2020

2020
[31]

Optimal robustness-consistency trade-offs for learning- augmentedonlinealgorithms.Advances in Neural Information Processing Systems, 33:8042– 8053, 2020

Alexander Wei and Fred Zhang. Optimal robustness-consistency trade-offs for learning- augmentedonlinealgorithms.Advances in Neural Information Processing Systems, 33:8042– 8053, 2020

2020
[32]

Strategyproof mechanism design for facility location games with weighted agents on a line.Journal of Combinatorial Optimization, 28:756–773, 2014

Qiang Zhang and Minming Li. Strategyproof mechanism design for facility location games with weighted agents on a line.Journal of Combinatorial Optimization, 28:756–773, 2014. 17

2014

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [28]

Optimality of the coordinate-wise median mechanism for strategyproof facility location in two dimensions

Goel Sumit and Hann-Caruthers Wade. Coordinate-wise median: Not bad, not bad, pretty good.arXiv preprint arXiv:2007.00903, 2020

work page Pith review arXiv 2007

[2] [24]

Range convexity, continuity, and strategy-proofness of voting schemes.Zeitschrift für Operations Research, 38:213–229, 1993

Hans Peters, Hans van der Stel, and Ton Storcken. Range convexity, continuity, and strategy-proofness of voting schemes.Zeitschrift für Operations Research, 38:213–229, 1993

work page 1993

[3] [1]

Learning- augmented mechanism design: Leveraging predictions for facility location

Priyank Agrawal, Eric Balkanski, Vasilis Gkatzelis, Tingting Ou, and Xizhi Tan. Learning- augmented mechanism design: Leveraging predictions for facility location. InProceedings of the 23rd ACM Conference on Economics and Computation, pages 497–528, 2022

work page 2022

[4] [2]

A survey of healthcare facility location.Computers & Operations Research, 79:223–263, 2017

Amir Ahmadi-Javid, Pardis Seyedi, and Siddhartha S Syam. A survey of healthcare facility location.Computers & Operations Research, 79:223–263, 2017. 15

work page 2017

[5] [3]

Secretary and online matching problems with machine learned advice.Advances in Neural Information Processing Systems, 33:7933–7944, 2020

Antonios Antoniadis, Themis Gouleakis, Pieter Kleer, and Pavel Kolev. Secretary and online matching problems with machine learned advice.Advances in Neural Information Processing Systems, 33:7933–7944, 2020

2020

[6] [4]

Randomized strategic facility location with predictions.Advances in Neural Information Processing Systems, 37:35639– 35664, 2024

Eric Balkanski, Vasilis Gkatzelis, and Golnoosh Shahkarami. Randomized strategic facility location with predictions.Advances in Neural Information Processing Systems, 37:35639– 35664, 2024

work page 2024

[7] [5]

Facility location optimization model for emergency humanitarian logistics.International Journal of Disaster Risk Reduc- tion, 24:485–498, 2017

Chawis Boonmee, Mikiharu Arimura, and Takumi Asada. Facility location optimization model for emergency humanitarian logistics.International Journal of Disaster Risk Reduc- tion, 24:485–498, 2017

work page 2017

[8] [6]

Cambridge University Press, 2016

Felix Brandt, Vincent Conitzer, Ulle Endriss, Jérôme Lang, and Ariel D Procaccia.Hand- book of computational social choice. Cambridge University Press, 2016

work page 2016