arXiv subjects

Max Dupré la Tour

Publications and source records attributed to Max Dupré la Tour.

16 recordsLinked to original sources

arXiv2026

Differentially Private Multicolor Discrepancy and Fair Division of Indivisible Goods

We study the fair division of indivisible goods under pure differential privacy, continuing the line of work initiated by Manurangsi and Suksompong. For $n$ agents with nonnegative additive utilities over $m$ goods and a fixed privacy parameter, we give an entry-private algorithm that, with high probability, achieves consensus envy-freeness up to $O(\sqrt n+\log^3 m)$ goods. This substantially improves the dependence on $n$ over the previous $O(n\log m)$ guarantee for ordinary envy-freeness, while providing the stronger consensus guarantee. A key ingredient is a private algorithm for multicolor discrepancy, which may be of independent interest. Our algorithm may require exponential time. We also obtain substantially stronger guarantees under additional structure: when all item values belong to a public alphabet of size $D$, we give a polynomial-time entry-private algorithm achieving ordinary envy-freeness up to $O(\operatorname{polylog}(mD))$ goods with high probability. Finally, we prove an $Ω(\log n)$ lower bound on the number of goods that must be removed to achieve ordinary envy-freeness under entry privacy, for sufficiently many goods, even with binary utilities.

cs.GT↗

arXiv2026

Subquadratic Subsidies for Nonnegative or Nonpositive Valuations

We study envy-freeness with subsidies for indivisible items beyond additive valuations. Assuming that every single-item marginal value lies in $[-1,1]$, we prove that a total subsidy of $O(n^{3/2}\sqrt{\log n})$ suffices to achieve envy-freeness among $n$ agents whenever all agents assign nonnegative values to every bundle or all assign nonpositive values to every bundle. These valuation classes include monotone goods and monotone chores, respectively, but do not require monotonicity. Our result establishes the first subquadratic total-subsidy bound for general monotone valuations that holds for every number of agents.

cs.GT↗

arXiv2026

Balanced Fair Division for Three Agents under General Valuations and Laminar Constraints

We study fair allocations of indivisible items under general set valuations. We prove that every instance with three agents and arbitrary real-valued valuations admits a balanced allocation that is envy-free up to one good and one chore (EF$1^c_g$). This directly implies balanced EF$1$ when each valuation is either monotone nondecreasing or monotone nonincreasing. We also show that balanced EF$1$ cannot be guaranteed without monotonicity: there exists an instance with three agents, nine items, and identical nonmonotone valuations that admits no balanced EF$1$ allocation. We then consider a common laminar matroid constraint. Whenever a complete feasible allocation exists, we prove that there is a complete feasible allocation that is balanced and envy-free up to two goods and two chores (EF$2^c_g$) for arbitrary valuations. The allocation can additionally be chosen so that the numbers of items from every laminar set assigned to the three agents differ by at most two. Most proofs in this paper were obtained using GPT-5.6. We subsequently verified the proofs for correctness and refined their exposition and arguments, also with the aid of GPT-5.6.

cs.GT↗

arXiv2026

From Cake-Cutting and Necklace-Splitting to Fair Division of Indivisible Items

We give an existential transfer framework for converting continuous fair division theorems into guarantees for indivisible items arranged on a path. This allows continuous envy-freeness and consensus results to translate directly into EF$k$-type guarantees for indivisible allocations. Combining this method with connected cake-cutting theorems, we obtain connected allocations satisfying envy-freeness up to one good and one chore for identical valuations and for arbitrary valuations when the number of agents is a prime power. Combining this method with the equicardinal necklace-splitting theorem of Jojić et al., we show that, for any prime-power number $r$ of bundles and $n$ arbitrary valuation functions, there exists an allocation in which every bundle is the union of at most $n$ intervals, and the bundles satisfy consensus up to $n$ goods and $n$ chores. This result is the first EF$k$-type guarantee for consensus fair division with non-additive valuations beyond the halving case. Envy-freeness constraints can be imposed simultaneously at the cost of one additional interval and one additional item in each guarantee. As a consequence, when the number of agents is a prime power, every instance with monotone valuations admits an EF$2$ allocation whose bundle sizes differ by at most two.

cs.GT↗

arXiv2026

Bad News for Couples: Bounds for Fair Division of Indivisible Items among Groups

We consider the problem of fairly allocating indivisible items to couples, where each couple consists of two agents with distinct additive valuations. We show that there exist binary instances with $n$ agents partitioned into $n/2$ couples for which envy-freeness up to $Ω(\sqrt{n})$ items cannot be guaranteed. More generally, in the group-allocation model with $n$ agents partitioned into $k$ groups, we construct binary instances for which envy-freeness up to $Ω(\sqrt{n-k})$ items cannot be guaranteed. This matches the $O(\sqrt{n})$ upper bound of Manurangsi and Suksompong in all regimes except when $n-k \ll n$, that is, when most agents form singleton groups and only a sublinear number of agents belong to groups of size at least two. This result is somewhat surprising, as that upper bound was conjectured not to be tight for instances consisting only of small groups, such as couples. We complement our lower bound with improved upper bounds for the remaining sparse regime. For prime-power $k$, we prove an upper bound of $O(\min\{\sqrt{(n-k)\log k},n-k\})$. For arbitrary $k$, the bound incurs an additional factor of $O(\frac{\log m}{\log k})$ where $m$ is the number of items. The result follows from a more general theorem that simultaneously guarantees approximate envy-freeness with respect to agents' subjective valuations and approximate equality with respect to multiple consensus valuations.

cs.GT↗

arXiv2026

Faster and Simpler Greedy Algorithm for $k$-Median and $k$-Means

Clustering problems such as $k$-means and $k$-median are staples of unsupervised learning, and many algorithmic techniques have been developed to tackle their numerous aspects. In this paper, we focus on the class of greedy approximation algorithm, that attracted less attention than local-search or primal-dual counterparts. In particular, we study the recursive greedy algorithm developed by Mettu and Plaxton [SIAM J. Comp 2003]. We provide a simplification of the algorithm, allowing for faster implementation, in graph metrics or in Euclidean space, where our algorithm matches or improves the state-of-the-art.

cs.DS↗

arXiv2026

Random Serial Dictatorship is $\sqrt{2}$-Envy-Free

We analyze the house allocation problem, in which a set of agents must be matched to a set of objects for which they have cardinal utilities. A central mechanism for this problem is random serial dictatorship (RSD), which has long served as a canonical subject of study due to its simplicity and the existence of exact characterizations by its properties. Despite this extensive understanding, a basic quantitative question about the fairness of this mechanism remains unresolved. Although RSD is often viewed as fair ex ante, surprisingly, it is not envy-free in expectation. We quantify its deviation from envy-freeness via the envy-ratio, which is the maximum over all instances and pairs of agents of the ratio between an agent's expected utility for another agent's random object and for its own random object. Prior work shows a factor-$\sqrt{2}\approx 1.414$ lower bound on the envy-ratio of RSD. Our headline result is a matching upper bound, showing that RSD is $\sqrt{2}$-envy-free in the house allocation problem. We further analyze the two natural extensions of RSD (the randomized round-robin mechanism and the iterated-RSD mechanism) to settings with unequal numbers of agents and objects and more general valuation classes. For additive valuations, this ratio increases to at least $1.5$ and at most $1.707$ for randomized round-robin, but remains exactly $\sqrt{2}$ for iterated-RSD. For submodular valuations, we prove constant-factor upper and lower bounds for both mechanisms, leaving only a small constant gap in both cases. For the more general classes of XOS and subadditive valuations, we present a tight analysis for both mechanisms, showing that the envy-ratio is unbounded in the number of agents. These results provide the first tight or nearly tight quantitative guarantees on the extent to which random serial dictatorship and its natural generalizations approximate envy-freeness.

cs.GT↗

arXiv2026

Tight Asymptotic Bounds for Fair Division With Externalities

We study the problem of allocating a set of indivisible items among agents whose preferences include externalities. Unlike the standard fair division model, agents may derive positive or negative utility not only from items allocated directly to them, but also from items allocated to other agents. Since exact envy-freeness cannot be guaranteed, prior work has focused on its relaxations. However, two central questions remained open: does there always exist an allocation that is envy-free up to one item (EF1), and if not, what is the optimal relaxation EF-$k$ that can always be attained? We settle both questions by deriving tight asymptotic bounds on the number of items sufficient to eliminate envy. We show that for any instance with $n$ agents, an allocation that is envy-free up to $O(\sqrt{n})$ items always exists and can be found in polynomial time, and we prove a matching $Ω(\sqrt{n})$ lower bound showing that this result is tight even for binary valuations, which rules out the existence of EF1 allocations when agents have externalities.

cs.GT↗

arXiv2025

On the hardness of recognizing graphs of small mim-width and its variants

The mim-width of a graph is a powerful structural parameter that, when bounded by a constant, allows several hard problems to be polynomial-time solvable - with a recent meta-theorem encompassing a large class of problems [SODA2023]. Since its introduction, several variants such as sim-width and omim-width were developed, along with a linear version of these parameters. It was recently shown that mim-width and all these variants all paraNP-hard, a consequence of the NP-hardness of distinguishing between graphs of linear mim-width at most 1211 and graphs of sim-width at least 1216 [ICALP2025]. The complexity of recognizing graphs of small width, particularly those close to $1$, remained open, despite their especially attractive algorithmic applications. In this work, we show that the width recognition problems remain NP-hard even on small widths. Specifically, after introducing the novel parameter Omim-width sandwiched between omim-width and mim-width, we show that: (1) deciding whether a graph has sim-width = 1, omim-width = 1, or Omin-width = 1 is NP-hard, and the same is true for their linear variants; (2) the problems of deciding whether mim-width $\leq$ 2 or linear mim-width $\leq$ 2 are both NP-hard. Interestingly, our reductions are relatively simple and are from the Unrooted Quartet Consistency problem, which is of great interest in computational biology but is not commonly used (if ever) in the theory of algorithms.

cs.DM↗

arXiv2025

Recognizing Leaf Powers and Pairwise Compatibility Graphs is NP-Complete

Leaf powers and pairwise compatibility graphs were introduced over twenty years ago as simplified graph models for phylogenetic trees. Despite significant research, several properties of these graph classes remain poorly understood. In this paper, we establish that the recognition problem for both classes is NP-complete. We extend this hardness result to a broader hierarchy of graph classes, including pairwise compatibility graphs and their generalizations, multi interval pairwise compatibility graphs.

math.CO↗

arXiv2024

Eliminating Majority Illusion is Easy

Majority Illusion is a phenomenon in social networks wherein the decision by the majority of the network is not the same as one's personal social circle's majority, leading to an incorrect perception of the majority in a large network. In this paper, we present polynomial-time algorithms which can eliminate majority illusion in a network by altering as few connections as possible. Additionally, we prove that the more general problem of ensuring all neighbourhoods in the network are at least a $p$-fraction of the majority is NP-hard for most values of $p$.

cs.MA↗

arXiv2024

$k$-Leaf Powers Cannot be Characterized by a Finite Set of Forbidden Induced Subgraphs for $k \geq 5$

A graph $G=(V,E)$ is a $k$-leaf power if there is a tree $T$ whose leaves are the vertices of $G$ with the property that a pair of leaves $u$ and $v$ induce an edge in $G$ if and only if they are distance at most $k$ apart in $T$. For $k\le 4$, it is known that there exists a finite set $F_k$ of graphs such that the class $L(k)$ of $k$-leaf power graphs is characterized as the set of strongly chordal graphs that do not contain any graph in $F_k$ as an induced subgraph. We prove no such characterization holds for $k\ge 5$. That is, for any $k\ge 5$, there is no finite set $F_k$ of graphs such that $L(k)$ is equivalent to the set of strongly chordal graphs that do not contain as an induced subgraph any graph in $F_k$.

math.CO↗

arXiv2024

Fully Dynamic k-Means Coreset in Near-Optimal Update Time

We study in this paper the problem of maintaining a solution to $k$-median and $k$-means clustering in a fully dynamic setting. To do so, we present an algorithm to efficiently maintain a coreset, a compressed version of the dataset, that allows easy computation of a clustering solution at query time. Our coreset algorithm has near-optimal update time of $\tilde O(k)$ in general metric spaces, which reduces to $\tilde O(d)$ in the Euclidean space $\mathbb{R}^d$. The query time is $O(k^2)$ in general metrics, and $O(kd)$ in $\mathbb{R}^d$. To maintain a constant-factor approximation for $k$-median and $k$-means clustering in Euclidean space, this directly leads to an algorithm update time $\tilde O(d)$, and query time $\tilde O(kd + k^2)$. To maintain a $O(polylog~k)$-approximation, the query time is reduced to $\tilde O(kd)$.

cs.DS↗

arXiv2024

Making Old Things New: A Unified Algorithm for Differentially Private Clustering

As a staple of data analysis and unsupervised learning, the problem of private clustering has been widely studied under various privacy models. Centralized differential privacy is the first of them, and the problem has also been studied for the local and the shuffle variation. In each case, the goal is to design an algorithm that computes privately a clustering, with the smallest possible error. The study of each variation gave rise to new algorithms: the landscape of private clustering algorithms is therefore quite intricate. In this paper, we show that a 20-year-old algorithm can be slightly modified to work for any of these models. This provides a unified picture: while matching almost all previously known results, it allows us to improve some of them and extend it to a new privacy model, the continual observation setting, where the input is changing over time and the algorithm must output a new solution at each time step.

cs.DS↗

arXiv2024

Gerrymandering Planar Graphs

We study the computational complexity of the map redistricting problem (gerrymandering). Mathematically, the electoral district designer (gerrymanderer) attempts to partition a weighted graph into $k$ connected components (districts) such that its candidate (party) wins as many districts as possible. Prior work has principally concerned the special cases where the graph is a path or a tree. Our focus concerns the realistic case where the graph is planar. We prove that the gerrymandering problem is solvable in polynomial time in $λ$-outerplanar graphs, when the number of candidates and $λ$ are constants and the vertex weights (voting weights) are polynomially bounded. In contrast, the problem is NP-complete in general planar graphs even with just two candidates. This motivates the study of approximation algorithms for gerrymandering planar graphs. However, when the number of candidates is large, we prove it is hard to distinguish between instances where the gerrymanderer cannot win a single district and instances where the gerrymanderer can win at least one district. This immediately implies that the redistricting problem is inapproximable in polynomial time in planar graphs, unless P=NP. This conclusion appears terminal for the design of good approximation algorithms -- but it is not. The inapproximability bound can be circumvented as it only applies when the maximum number of districts the gerrymanderer can win is extremely small, say one. Indeed, for a fixed number of candidates, our main result is that there is a constant factor approximation algorithm for redistricting unweighted planar graphs, provided the optimal value is a large enough constant.

cs.GT↗

arXiv2023

Differential Privacy for Clustering Under Continual Observation

We consider the problem of clustering privately a dataset in $\mathbb{R}^d$ that undergoes both insertion and deletion of points. Specifically, we give an $\varepsilon$-differentially private clustering mechanism for the $k$-means objective under continual observation. This is the first approximation algorithm for that problem with an additive error that depends only logarithmically in the number $T$ of updates. The multiplicative error is almost the same as non privately. To do so we show how to perform dimension reduction under continual observation and combine it with a differentially private greedy approximation algorithm for $k$-means. We also partially extend our results to the $k$-median problem.

cs.DS↗