Hypergraphs and Decision Trees (Abstract).

Andrew Chi-Chih Yao

DOI: https://doi.org/10.1007/3-540-62559-3_1

1996-01-01

Abstract:We survey some recent results on a space decomposition problem that has a close relationship to the size of algebraic decision trees. These results are obtained by establishing connections between the decomposition problem and some extremal questions in hypergraphs.

What problem does this paper attempt to address?

Hypergraphs And Decision Trees

Ac Yao

DOI: https://doi.org/10.1007/3-540-62559-3_1

1997-01-01

Abstract:We survey some recent results on a space decomposition problem that has a close relationship to the size of algebraic decision trees. These results are obtained by establishing connections between the decomposition problem and some extremal questions in hypergraphs.Let d, n less-than-or-equal-to 1 be integers. A d-elementary cell is a set D in-set-of R-n defined by a finite set of constraints f(i)(x) = 0, g(j)(x) greater-than 0 where f(i), g(j) are polynomials of degrees not exceeding d. For any set S in-the-set-of R-n, let K-d(S) be the smallest number of disjoint d-elementary cells that S can be decomposed into. It is easy to see that any degree-d (ternary) algebraic decision tree for solving the membership question of S must have size no less than K-d(S). Thus, any lower bound to K-d(S) yields also a lower bound the size complexity for the corresponding membership problem.Let A(n) = {x(1), x(2), ... x(n)/ x(i) less-than-or-equal-to 0}. A well-known result of Rabin states that any algebraic decision tree for the membership question of A(n) must have height at least n. In this talk we discuss a recent result by Grigoriev, Karpinski and Yao [GKY], which gives an exponential lower bound to K-d(A(n)) for any fixed d, and hence to the size of any fixed degree (ternary) algebraic decision tree for solving this problem. The proof utilized a new connection between K-d(A(n)) and the maximum number of minimal cutsets for any rank-d hypergraph on n vertices. We also discuss an improved lower bound by Wigderson and Yao [WY]. Open questions are presented.
Fast hyperboloid decision tree algorithms

Philippe Chlenski,Ethan Turok,Antonio Moretti,Itsik Pe'er

DOI: https://doi.org/10.48550/arXiv.2310.13841

IF: 5.414

2023-10-20

Machine Learning

Abstract:Hyperbolic geometry is gaining traction in machine learning for its effectiveness at capturing hierarchical structures in real-world data. Hyperbolic spaces, where neighborhoods grow exponentially, offer substantial advantages and consistently deliver state-of-the-art results across diverse applications. However, hyperbolic classifiers often grapple with computational challenges. Methods reliant on Riemannian optimization frequently exhibit sluggishness, stemming from the increased computational demands of operations on Riemannian manifolds. In response to these challenges, we present hyperDT, a novel extension of decision tree algorithms into hyperbolic space. Crucially, hyperDT eliminates the need for computationally intensive Riemannian optimization, numerically unstable exponential and logarithmic maps, or pairwise comparisons between points by leveraging inner products to adapt Euclidean decision tree algorithms to hyperbolic space. Our approach is conceptually straightforward and maintains constant-time decision complexity while mitigating the scalability issues inherent in high-dimensional Euclidean spaces. Building upon hyperDT we introduce hyperRF, a hyperbolic random forest model. Extensive benchmarking across diverse datasets underscores the superior performance of these models, providing a swift, precise, accurate, and user-friendly toolkit for hyperbolic data analysis.
Open Problems in (Hyper)Graph Decomposition

Deepak Ajwani,Rob H. Bisseling,Katrin Casel,Ümit V. Çatalyürek,Cédric Chevalier,Florian Chudigiewitsch,Marcelo Fonseca Faraj,Michael Fellows,Lars Gottesbüren,Tobias Heuer,George Karypis,Kamer Kaya,Jakub Lacki,Johannes Langguth,Xiaoye Sherry Li,Ruben Mayer,Johannes Meintrup,Yosuke Mizutani,François Pellegrini,Fabrizio Petrini,Frances Rosamond,Ilya Safro,Sebastian Schlag,Christian Schulz,Roohani Sharma,Darren Strash,Blair D. Sullivan,Bora Uçar,Albert-Jan Yzelman

2023-10-18

Abstract:Large networks are useful in a wide range of applications. Sometimes problem instances are composed of billions of entities. Decomposing and analyzing these structures helps us gain new insights about our surroundings. Even if the final application concerns a different problem (such as traversal, finding paths, trees, and flows), decomposing large graphs is often an important subproblem for complexity reduction or parallelization. This report is a summary of discussions that happened at Dagstuhl seminar 23331 on "Recent Trends in Graph Decomposition" and presents currently open problems and future directions in the area of (hyper)graph decomposition.

Data Structures and Algorithms
Algebraic Decision Trees and Euler Characteristics

Andrew Chi-Chih Yao

DOI: https://doi.org/10.1109/sfcs.1992.267765

1992-01-01

Abstract:For any set S contained in R/sup n/, let chi (S) denote its Euler characteristic. The author shows that any algebraic computation tree or fixed-degree algebraic decision tree must have height Omega (log mod chi (S) mod )for deciding the membership question of a compact semi-algebraic set S. This extends a result by A. Bjorner, L. Lovasz and A. Yao where it was shown that any linear decision tree for deciding the membership question of a closed polyhedron S must have height greater than or equal to log/sub 3/ mod chi (S) mod .
Lower bounds for algebraic decision trees

J. Michael Steele,Andrew Chi-chih Yao

DOI: https://doi.org/10.1016/0196-6774(82)90002-5

1982-01-01

Journal of Algorithms

Abstract:A topological method is given for obtaining lower bounds for the height of algebraic decision trees. The method is applied to the knapsack problem where an Ω(n2) bound is obtained for trees with bounded-degree polynomial tests, thus extending the Dobkin-Lipton result for linear trees. Applications to the convex hull problem and the distinct element problem are also indicated. Some open problems are discussed.
An Exponential Lower Bound on the Size of Algebraic Decision Trees for MAX

D. Grigoriev,M. Karpinski,A.C. Yao

DOI: https://doi.org/10.1007/s000370050010

1998-01-01

Computational Complexity

Abstract:We prove an exponential lower bound on the size of any fixed degree algebraic decision tree for solving MAX, the problem of finding the maximum of n real numbers. This complements the n— 1 lower bound of [Rabin (1972)] on the depth of algebraic decision trees for this problem. The proof in fact gives an exponential lower bound on the size of the polyhedral decision problem MAX= for testing whether the j-th number is the maximum among a list of n real numbers. Previously, except for linear decision trees, no nontrivial lower bounds on the size of algebraic decision trees for any familiar problems are known. We also establish an interesting connection between our lower bound and the maximum number of minimal cutsets for any rank-d hypergraph on n vertices.
Tree Projections and Structural Decomposition Methods: Minimality and Game-Theoretic Characterization

Gianluigi Greco,Francesco Scarcello

DOI: https://doi.org/10.48550/arXiv.1212.2314

2012-12-11

Abstract:Tree projections provide a mathematical framework that encompasses all the various (purely) structural decomposition methods that have been proposed in the literature to single out classes of nearly-acyclic (hyper)graphs, such as the tree decomposition method, which is the most powerful decomposition method on graphs, and the (generalized) hypertree decomposition method, which is its natural counterpart on arbitrary hypergraphs. The paper analyzes this framework, by focusing in particular on "minimal" tree projections, that is, on tree projections without useless redundancies. First, it is shown that minimal tree projections enjoy a number of properties that are usually required for normal form decompositions in various structural decomposition methods. In particular, they enjoy the same kind of connection properties as (minimal) tree decompositions of graphs, with the result being tight in the light of the negative answer that is provided to the open question about whether they enjoy a slightly stronger notion of connection property, defined to speed-up the computation of hypertree decompositions. Second, it is shown that tree projections admit a natural game-theoretic characterization in terms of the Captain and Robber game. In this game, as for the Robber and Cops game characterizing tree decompositions, the existence of winning strategies implies the existence of monotone ones. As a special case, the Captain and Robber game can be used to characterize the generalized hypertree decomposition method, where such a game-theoretic characterization was missing and asked for. Besides their theoretical interest, these results have immediate algorithmic applications both for the general setting and for structural decomposition methods that can be recast in terms of tree projections.

Discrete Mathematics,Artificial Intelligence
Flowers in graph-like spaces

Ann-Kathrin Elm,Hendrik Heine

DOI: https://doi.org/10.48550/arXiv.2312.13825

2023-12-21

Combinatorics

Abstract:One perspective on tree decompositions is that they display (low-order) separations of the underlying graph or matroid. The separations displayed by a tree decomposition are necessarily nested. In 2013, Clark and Whittle proved the existence of tree decompositions with flowers added in which, up to a natural equivalence, all low-order separations are displayed. An important step in that proof is to show that flowers can be extended to maximal flowers. In this paper we generalise the notion of a flower to pseudoflowers in graph-like spaces and show for our generalisation, flowers can be extended to maximal flowers.
Realizing trees of hypergraphs in thin sets

Allan Greenleaf,Alex Iosevich,Krystal Taylor

2024-01-22

Abstract:Let $\phi(x,y)$ be a continuous function, smooth away from the diagonal, such that, for some $\alpha>0$, the associated generalized Radon transforms \begin{equation} \label{Radon} R_t^{\phi}f(x)=\int_{\phi(x,y)=t} f(y) \psi(y) d\sigma_{x,t}(y) \end{equation} map $L^2({\mathbb R}^d) \to H^{\alpha}({\mathbb R}^d)$ for all $t>0$. Let $E$ be a compact subset of ${\mathbb R}^d$ for some $d \ge 2$, and suppose that the Hausdorff dimension of $E$ is $>d-\alpha$. Then any tree graph $T$ on $k+1$ ($k \ge 1$) vertices is realizable in $E$, in the sense that there exist distinct $x^1, x^2, \dots, x^{k+1} \in E$ and $t>0$ such that the $\phi$-distance $\phi(x^i, x^j)$ is equal to $t$ for all pairs $(i,j)$ corresponding to the edges of the graph $T$.

Classical Analysis and ODEs,Combinatorics,Metric Geometry
Large monochromatic components in expansive hypergraphs

Deepak Bal,Louis DeBiasio

DOI: https://doi.org/10.1017/s096354832400004x

2024-03-14

Abstract:A result of Gyárfás [12] exactly determines the size of a largest monochromatic component in an arbitrary -colouring of the complete -uniform hypergraph when and . We prove a result which says that if one replaces in Gyárfás' theorem by any 'expansive' -uniform hypergraph on vertices (that is, a -uniform hypergraph on vertices in which for all disjoint sets with for all ), then one gets a largest monochromatic component of essentially the same size (within a small error term depending on and ). As corollaries we recover a number of known results about large monochromatic components in random hypergraphs and random Steiner triple systems, often with drastically improved bounds on the error terms. Gyárfás' result is equivalent to the dual problem of determining the smallest possible maximum degree of an arbitrary -partite -uniform hypergraph with edges in which every set of edges has a common intersection. In this language, our result says that if one replaces the condition that every set of edges has a common intersection with the condition that for every collection of disjoint sets with , there exists such that , then the smallest possible maximum degree of is essentially the same (within a small error term depending on and ). We prove our results in this dual setting.
The HyperTrac Project: Recent Progress and Future Research Directions on Hypergraph Decompositions

Georg Gottlob,Matthias Lanzinger,Davide Mario Longo,Cem Okulmus,Reinhard Pichler

DOI: https://doi.org/10.48550/arXiv.2012.14762

2020-12-29

Abstract:Constraint Satisfaction Problems (CSPs) play a central role in many applications in Artificial Intelligence and Operations Research. In general, solving CSPs is NP-complete. The structure of CSPs is best described by hypergraphs. Therefore, various forms of hypergraph decompositions have been proposed in the literature to identify tractable fragments of CSPs. However, also the computation of a concrete hypergraph decomposition is a challenging task in itself. In this paper, we report on recent progress in the study of hypergraph decompositions and we outline several directions for future research.

Artificial Intelligence
Complexity Analysis of Generalized and Fractional Hypertree Decompositions

Georg Gottlob,Matthias Lanzinger,Reinhard Pichler,Igor Razgon

DOI: https://doi.org/10.1145/3457374

2020-03-05

Abstract:Hypertree decompositions (HDs), as well as the more powerful generalized hypertree decompositions (GHDs), and the yet more general fractional hypertree decompositions (FHDs) are hypergraph decomposition methods successfully used for answering conjunctive queries and for solving constraint satisfaction problems. Every hypergraph $H$ has a width relative to each of these methods: its hypertree width $hw(H)$, its generalized hypertree width $ghw(H)$, and its fractional hypertree width $fhw(H)$, respectively. It is known that $hw(H)\leq k$ can be checked in polynomial time for fixed $k$, while checking $ghw(H)\leq k$ is NP-complete for $k \geq 3$. The complexity of checking $fhw(H)\leq k$ for a fixed $k$ has been open for over a decade. We settle this open problem by showing that checking $fhw(H)\leq k$ is NP-complete, even for $k=2$. The same construction allows us to prove also the NP-completeness of checking $ghw(H)\leq k$ for $k=2$. After that, we identify meaningful restrictions which make checking for bounded $ghw$ or $fhw$ tractable or allow for an efficient approximation of the $fhw$.

Computational Complexity
Diversities and the Geometry of Hypergraphs

David Bryant,Paul F. Tupper

DOI: https://doi.org/10.48550/arXiv.1312.5408

2014-04-18

Abstract:The embedding of finite metrics in $\ell_1$ has become a fundamental tool for both combinatorial optimization and large-scale data analysis. One important application is to network flow problems in which there is close relation between max-flow min-cut theorems and the minimal distortion embeddings of metrics into $\ell_1$. Here we show that this theory can be generalized considerably to encompass Steiner tree packing problems in both graphs and hypergraphs. Instead of the theory of $\ell_1$ metrics and minimal distortion embeddings, the parallel is the theory of diversities recently introduced by Bryant and Tupper, and the corresponding theory of $\ell_1$ diversities and embeddings which we develop here.

Metric Geometry,Data Structures and Algorithms,Combinatorics
Decomposition horizons and a characterization of stable hereditary classes of graphs

Samuel Braunfeld,Jaroslav Nešetřil,Patrice Ossona de Mendez,Sebastian Siebertz

2024-01-19

Abstract:The notions of bounded-size and quasibounded-size decompositions with bounded treedepth base classes are central to the structural theory of graph sparsity introduced by two of the authors years ago, and provide a characterization of both classes with bounded expansions and nowhere dense classes. In this paper, we first prove that the model theoretic notions of dependence and stability are, for hereditary classes of graphs, compatible with quasibounded-size decompositions, in the following sense: every hereditary class with quasibounded-size decompositions with dependent (resp.\ stable) base classes is itself dependent (resp.\ stable). This result is obtained in a more general study of ``decomposition horizons'', which are class properties compatible with quasibounded-size decompositions. We deduce that hereditary classes with quasibounded-size decompositions with bounded shrubdepth base classes are stable. In the second part of the paper, we prove the converse. Thus, we characterize stable hereditary classes of graphs as those hereditary classes that admit quasibounded-size decompositions with bounded shrubdepth base classes. This result is obtained by proving that every hereditary stable class of graphs admits almost nowhere dense quasi-bush representations, thus answering positively a conjecture of Dreier et al. These results have several consequences. For example, we show that every graph $G$ in a stable, hereditary class of graphs $\mathscr C$ has a clique or a stable set of size $\Omega_{\mathscr C,\epsilon}(|G|^{1/2-\epsilon})$, for every $\epsilon>0$, which is tight in the sense that it cannot be improved to $\Omega_{\mathscr C}(|G|^{1/2})$.

Discrete Mathematics,Logic in Computer Science,Combinatorics,Logic
Horospherical Decision Boundaries for Large Margin Classification in Hyperbolic Space

Xiran Fan,Chun-Hao Yang,Baba C Vemuri

Abstract:Hyperbolic spaces have been quite popular in the recent past for representing hierarchically organized data. Further, several classification algorithms for data in these spaces have been proposed in the literature. These algorithms mainly use either hyperplanes or geodesics for decision boundaries in a large margin classifiers setting leading to a non-convex optimization problem. In this paper, we propose a novel large margin classifier based on horospherical decision boundaries that leads to a geodesically convex optimization problem that can be optimized using any Riemannian gradient descent technique guaranteeing a globally optimal solution. We present several experiments depicting the competitive performance of our classifier in comparison to SOTA.
Near-optimal Linear Decision Trees for k-SUM and Related Problems

Daniel M. Kane,Shachar Lovett,Shay Moran

DOI: https://doi.org/10.1145/3285953

IF: 2.269

2019-06-30

Journal of the ACM

Abstract:We construct near-optimal linear decision trees for a variety of decision problems in combinatorics and discrete geometry. For example, for any constant k , we construct linear decision trees that solve the k -SUM problem on n elements using O ( n log 2 n ) linear queries. Moreover, the queries we use are comparison queries, which compare the sums of two k -subsets; when viewed as linear queries, comparison queries are 2 k -sparse and have only { −1,0,1} coefficients. We give similar constructions for sorting sumsets A+B and for solving the SUBSET-SUM problem, both with optimal number of queries, up to poly-logarithmic terms. Our constructions are based on the notion of “inference dimension,” recently introduced by the authors in the context of active classification with comparison queries. This can be viewed as another contribution to the fruitful link between machine learning and discrete geometry, which goes back to the discovery of the VC dimension.

computer science, information systems, theory & methods, software engineering, hardware & architecture
Matroid-reachability-based decomposition into arborescences

Florian Hörsch,Benjamin Peyrille,Zoltán Szigeti

2024-05-06

Abstract:The problem of matroid-reachability-based packing of arborescences was solved by Király. Here we solve the corresponding decomposition problem that turns out to be more complicated. The result is obtained from the solution of the more general problem of matroid-reachability-based $(\ell,\ell')$-limited packing of arborescences where we are given a lower bound $\ell$ and an upper bound $\ell'$ on the total number of arborescences in the packing. The problem is considered for branchings and in directed hypergraphs as well.

Combinatorics,Discrete Mathematics
Bounded-Diameter Tree-Decompositions

Eli Berger,Paul Seymour

DOI: https://doi.org/10.1007/s00493-024-00088-1

2024-04-09

COMBINATORICA

Abstract:When does a graph admit a tree-decomposition in which every bag has small diameter? For finite graphs, this is a property of interest in algorithmic graph theory, where it is called having bounded "tree-length". We will show that this is equivalent to being "boundedly quasi-isometric to a tree", which for infinite graphs is a much-studied property from metric geometry. One object of this paper is to tie these two areas together. We will prove that there is a tree-decomposition in which each bag has small diameter, if and only if there is a map from V ( G ) into the vertex set of a tree T , such that for all , the distances differ by at most a constant. A necessary condition for admitting such a tree-decomposition is that there is no long geodesic cycle, and for graphs of bounded tree-width, Diestel and Müller showed that this is also sufficient. But it is not sufficient in general, even qualitatively, because there are graphs in which every geodesic cycle has length at most three, and yet every tree-decomposition has a bag with large diameter. There is a more general necessary condition, however. A "geodesic loaded cycle" in G is a pair ( C , F ), where C is a cycle of G and , such that for every pair u , v of vertices of C , one of the paths of C between u , v contains at most F -edges, where is the distance between u , v in G . We will show that a (possibly infinite) graph G admits a tree-decomposition in which every bag has small diameter, if and only if | F | is small for every geodesic loaded cycle ( C , F ). Our proof is an extension of an algorithm to approximate tree-length in finite graphs by Dourisboure and Gavoille. In metric geometry, there is a similar theorem that characterizes when a graph is quasi-isometric to a tree, "Manning's bottleneck criterion". The goal of this paper is to tie all these concepts together, and add a few more related ideas. For instance, we prove a conjecture of Rose McCarty, that G admits a tree-decomposition in which every bag has small diameter, if and only if for all vertices u , v , w of G , some ball of small radius meets every path joining two of u , v , w .

mathematics
Tree decompositions and many-sided separations

Tara Abrishami

DOI: https://doi.org/10.48550/arXiv.2207.10778

2022-07-22

Abstract:A separation of a graph $G$ is a partition $(A_1, A_2, C)$ of $V(G)$ such that $A_1$ is anticomplete to $A_2$. A classic result from Robertson and Seymour's Graph Minors Project states that there is a correspondence between tree decompositions and laminar collections of separations. A many-sided separation of a graph $G$ is a partition $(A_1, \ldots, A_k, C)$ of $V(G)$ such that $A_i$ is anticomplete to $A_j$ for all $1 \leq i < j \leq k$. In this note, we show a correspondence between tree decompositions with a certain parity property, called deciduous tree decompositions, and laminar collections of many-sided separations.

Combinatorics
Representation Tradeoffs for Hyperbolic Embeddings

Christopher De Sa,Albert Gu,Christopher Ré,Frederic Sala

DOI: https://doi.org/10.48550/arXiv.1804.03329

2018-04-24

Abstract:Hyperbolic embeddings offer excellent quality with few dimensions when embedding hierarchical data structures like synonym or type hierarchies. Given a tree, we give a combinatorial construction that embeds the tree in hyperbolic space with arbitrarily low distortion without using optimization. On WordNet, our combinatorial embedding obtains a mean-average-precision of 0.989 with only two dimensions, while Nickel et al.'s recent construction obtains 0.87 using 200 dimensions. We provide upper and lower bounds that allow us to characterize the precision-dimensionality tradeoff inherent in any hyperbolic embedding. To embed general metric spaces, we propose a hyperbolic generalization of multidimensional scaling (h-MDS). We show how to perform exact recovery of hyperbolic points from distances, provide a perturbation analysis, and give a recovery result that allows us to reduce dimensionality. The h-MDS approach offers consistently low distortion even with few dimensions across several datasets. Finally, we extract lessons from the algorithms and theory above to design a PyTorch-based implementation that can handle incomplete information and is scalable.

Machine Learning

Hypergraphs and Decision Trees (Abstract).

Hypergraphs And Decision Trees

Fast hyperboloid decision tree algorithms

Open Problems in (Hyper)Graph Decomposition

Algebraic Decision Trees and Euler Characteristics

Lower bounds for algebraic decision trees

An Exponential Lower Bound on the Size of Algebraic Decision Trees for MAX

Tree Projections and Structural Decomposition Methods: Minimality and Game-Theoretic Characterization

Flowers in graph-like spaces

Realizing trees of hypergraphs in thin sets

Large monochromatic components in expansive hypergraphs

The HyperTrac Project: Recent Progress and Future Research Directions on Hypergraph Decompositions

Complexity Analysis of Generalized and Fractional Hypertree Decompositions

Diversities and the Geometry of Hypergraphs

Decomposition horizons and a characterization of stable hereditary classes of graphs

Horospherical Decision Boundaries for Large Margin Classification in Hyperbolic Space

Near-optimal Linear Decision Trees for k-SUM and Related Problems

Matroid-reachability-based decomposition into arborescences

Bounded-Diameter Tree-Decompositions

Tree decompositions and many-sided separations

Representation Tradeoffs for Hyperbolic Embeddings