arXiv subjects

Suman K. Bera

Publications and source records attributed to Suman K. Bera.

11 recordsLinked to original sources

arXiv2022

DeMEtRIS: Counting (near)-Cliques by Crawling

We study the problem of approximately counting cliques and near cliques in a graph, where the access to the graph is only available through crawling its vertices; thus typically seeing only a small portion of it. This model, known as the random walk model or the neighborhood query model has been introduced recently and captures real-life scenarios in which the entire graph is too massive to be stored as a whole or be scanned entirely and sampling vertices independently is non-trivial in it. We introduce DeMEtRIS: Dense Motif Estimation through Random Incident Sampling. This method provides a scalable algorithm for clique and near clique counting in the random walk model. We prove the correctness of our algorithm through rigorous mathematical analysis and extensive experiments. Both our theoretical results and our experiments show that DeMEtRIS obtains a high precision estimation by only crawling a sub-linear portion on vertices, thus we demonstrate a significant improvement over previously known results.

cs.DS↗

arXiv2022

A New Dynamic Algorithm for Densest Subhypergraphs

Computing a dense subgraph is a fundamental problem in graph mining, with a diverse set of applications ranging from electronic commerce to community detection in social networks. In many of these applications, the underlying context is better modelled as a weighted hypergraph that keeps evolving with time. This motivates the problem of maintaining the densest subhypergraph of a weighted hypergraph in a {\em dynamic setting}, where the input keeps changing via a sequence of updates (hyperedge insertions/deletions). Previously, the only known algorithm for this problem was due to Hu et al. [HWC17]. This algorithm worked only on unweighted hypergraphs, and had an approximation ratio of $(1+ε)r^2$ and an update time of $O(\text{poly} (r, \log n))$, where $r$ denotes the maximum rank of the input across all the updates. We obtain a new algorithm for this problem, which works even when the input hypergraph is weighted. Our algorithm has a significantly improved (near-optimal) approximation ratio of $(1+ε)$ that is independent of $r$, and a similar update time of $O(\text{poly} (r, \log n))$. It is the first $(1+ε)$-approximation algorithm even for the special case of weighted simple graphs. To complement our theoretical analysis, we perform experiments with our dynamic algorithm on large-scale, real-world data-sets. Our algorithm significantly outperforms the state of the art [HWC17] both in terms of accuracy and efficiency.

cs.DS↗

arXiv2021

Counting Subgraphs in Degenerate Graphs

We consider the problem of counting the number of copies of a fixed graph $H$ within an input graph $G$. This is one of the most well-studied algorithmic graph problems, with many theoretical and practical applications. We focus on solving this problem when the input $G$ has bounded degeneracy. This is a rich family of graphs, containing all graphs without a fixed minor (e.g. planar graphs), as well as graphs generated by various random processes (e.g. preferential attachment graphs). We say that $H$ is easy if there is a linear-time algorithm for counting the number of copies of $H$ in an input $G$ of bounded degeneracy. A seminal result of Chiba and Nishizeki from '85 states that every $H$ on at most 4 vertices is easy. Bera, Pashanasangi, and Seshadhri recently extended this to all $H$ on 5 vertices, and further proved that for every $k > 5$ there is a $k$-vertex $H$ which is not easy. They left open the natural problem of characterizing all easy graphs $H$. Bressan has recently introduced a framework for counting subgraphs in degenerate graphs, from which one can extract a sufficient condition for a graph $H$ to be easy. Here we show that this sufficient condition is also necessary, thus fully answering the Bera--Pashanasangi--Seshadhri problem. We further resolve two closely related problems; namely characterizing the graphs that are easy with respect to counting induced copies, and with respect to counting homomorphisms.

cs.CC↗

arXiv2020

Near-Linear Time Homomorphism Counting in Bounded Degeneracy Graphs: The Barrier of Long Induced Cycles

Counting homomorphisms of a constant sized pattern graph $H$ in an input graph $G$ is a fundamental computational problem. There is a rich history of studying the complexity of this problem, under various constraints on the input $G$ and the pattern $H$. Given the significance of this problem and the large sizes of modern inputs, we investigate when near-linear time algorithms are possible. We focus on the case when the input graph has bounded degeneracy, a commonly studied and practically relevant class for homomorphism counting. It is known from previous work that for certain classes of $H$, $H$-homomorphisms can be counted exactly in near-linear time in bounded degeneracy graphs. Can we precisely characterize the patterns $H$ for which near-linear time algorithms are possible? We completely resolve this problem, discovering a clean dichotomy using fine-grained complexity. Let $m$ denote the number of edges in $G$. We prove the following: if the largest induced cycle in $H$ has length at most $5$, then there is an $O(m\log m)$ algorithm for counting $H$-homomorphisms in bounded degeneracy graphs. If the largest induced cycle in $H$ has length at least $6$, then (assuming standard fine-grained complexity conjectures) there is a constant $γ> 0$, such that there is no $o(m^{1+γ})$ time algorithm for counting $H$-homomorphisms.

cs.DS↗

arXiv2020

How to Count Triangles, without Seeing the Whole Graph

Triangle counting is a fundamental problem in the analysis of large graphs. There is a rich body of work on this problem, in varying streaming and distributed models, yet all these algorithms require reading the whole input graph. In many scenarios, we do not have access to the whole graph, and can only sample a small portion of the graph (typically through crawling). In such a setting, how can we accurately estimate the triangle count of the graph? We formally study triangle counting in the {\em random walk} access model introduced by Dasgupta et al (WWW '14) and Chierichetti et al (WWW '16). We have access to an arbitrary seed vertex of the graph, and can only perform random walks. This model is restrictive in access and captures the challenges of collecting real-world graphs. Even sampling a uniform random vertex is a hard task in this model. Despite these challenges, we design a provable and practical algorithm, TETRIS, for triangle counting in this model. TETRIS is the first provably sublinear algorithm (for most natural parameter settings) that approximates the triangle count in the random walk model, for graphs with low mixing time. Our result builds on recent advances in the theory of sublinear algorithms. The final sample built by TETRIS is a careful mix of random walks and degree-biased sampling of neighborhoods. Empirically, TETRIS accurately counts triangles on a variety of large graphs, getting estimates within 5\% relative error by looking at 3\% of the number of edges.

cs.DS↗

arXiv2020

Distributional Individual Fairness in Clustering

In this paper, we initiate the study of fair clustering that ensures distributional similarity among similar individuals. In response to improving fairness in machine learning, recent papers have investigated fairness in clustering algorithms and have focused on the paradigm of statistical parity/group fairness. These efforts attempt to minimize bias against some protected groups in the population. However, to the best of our knowledge, the alternative viewpoint of individual fairness, introduced by Dwork et al. (ITCS 2012) in the context of classification, has not been considered for clustering so far. Similar to Dwork et al., we adopt the individual fairness notion which mandates that similar individuals should be treated similarly for clustering problems. We use the notion of $f$-divergence as a measure of statistical similarity that significantly generalizes the ones used by Dwork et al. We introduce a framework for assigning individuals, embedded in a metric space, to probability distributions over a bounded number of cluster centers. The objective is to ensure (a) low cost of clustering in expectation and (b) individuals that are close to each other in a given fairness space are mapped to statistically similar distributions. We provide an algorithm for clustering with $p$-norm objective ($k$-center, $k$-means are special cases) and individual fairness constraints with provable approximation guarantee. We extend this framework to include both group fairness and individual fairness inside the protected groups. Finally, we observe conditions under which individual fairness implies group fairness. We present extensive experimental evidence that justifies the effectiveness of our approach.

cs.LG↗

arXiv2020

How the Degeneracy Helps for Triangle Counting in Graph Streams

We revisit the well-studied problem of triangle count estimation in graph streams. Given a graph represented as a stream of $m$ edges, our aim is to compute a $(1\pm\varepsilon)$-approximation to the triangle count $T$, using a small space algorithm. For arbitrary order and a constant number of passes, the space complexity is known to be essentially $Θ(\min(m^{3/2}/T, m/\sqrt{T}))$ (McGregor et al., PODS 2016, Bera et al., STACS 2017). We give a (constant pass, arbitrary order) streaming algorithm that can circumvent this lower bound for \emph{low degeneracy graphs}. The degeneracy, $κ$, is a nuanced measure of density, and the class of constant degeneracy graphs is immensely rich (containing planar graphs, minor-closed families, and preferential attachment graphs). We design a streaming algorithm with space complexity $\widetilde{O}(mκ/T)$. For constant degeneracy graphs, this bound is $\widetilde{O}(m/T)$, which is significantly smaller than both $m^{3/2}/T$ and $m/\sqrt{T}$. We complement our algorithmic result with a nearly matching lower bound of $Ω(mκ/T)$.

cs.DS↗

arXiv2019

Linear Time Subgraph Counting, Graph Degeneracy, and the Chasm at Size Six

We consider the problem of counting all $k$-vertex subgraphs in an input graph, for any constant $k$. This problem (denoted sub-cnt$_k$) has been studied extensively in both theory and practice. In a classic result, Chiba and Nishizeki (SICOMP 85) gave linear time algorithms for clique and 4-cycle counting for bounded degeneracy graphs. This is a rich class of sparse graphs that contains, for example, all minor-free families and preferential attachment graphs. The techniques from this result have inspired a number of recent practical algorithms for sub-cnt$_k$. Towards a better understanding of the limits of these techniques, we ask: for what values of $k$ can sub-cnt$_k$ be solved in linear time? We discover a chasm at $k=6$. Specifically, we prove that for $k < 6$, sub-cnt$_k$ can be solved in linear time. Assuming a standard conjecture in fine-grained complexity, we prove that for all $k \geq 6$, sub-cnt$_k$ cannot be solved even in near-linear time.

cs.DS↗

arXiv2019

Fair Algorithms for Clustering

We study the problem of finding low-cost Fair Clusterings in data where each data point may belong to many protected groups. Our work significantly generalizes the seminal work of Chierichetti et.al. (NIPS 2017) as follows. - We allow the user to specify the parameters that define fair representation. More precisely, these parameters define the maximum over- and minimum under-representation of any group in any cluster. - Our clustering algorithm works on any $\ell_p$-norm objective (e.g. $k$-means, $k$-median, and $k$-center). Indeed, our algorithm transforms any vanilla clustering solution into a fair one incurring only a slight loss in quality. - Our algorithm also allows individuals to lie in multiple protected groups. In other words, we do not need the protected groups to partition the data and we can maintain fairness across different groups simultaneously. Our experiments show that on established data sets, our algorithm performs much better in practice than what our theoretical results suggest.

cs.DS↗

arXiv2019

Graph Coloring via Degeneracy in Streaming and Other Space-Conscious Models

We study the problem of coloring a given graph using a small number of colors in several well-established models of computation for big data. These include the data streaming model, the general graph query model, the massively parallel computation (MPC) model, and the CONGESTED-CLIQUE and the LOCAL models of distributed computation. On the one hand, we give algorithms with sublinear complexity, for the appropriate notion of complexity in each of these models. Our algorithms color a graph $G$ using about $κ(G)$ colors, where $κ(G)$ is the degeneracy of $G$: this parameter is closely related to the arboricity $α(G)$. As a function of $κ(G)$ alone, our results are close to best possible, since the optimal number of colors is $κ(G)+1$. On the other hand, we establish certain lower bounds indicating that sublinear algorithms probably cannot go much further. In particular, we prove that any randomized coloring algorithm that uses $κ(G)+1$ many colors, would require $Ω(n^2)$ storage in the one pass streaming model, and $Ω(n^2)$ many queries in the general graph query model, where $n$ is the number of vertices in the graph. These lower bounds hold even when the value of $κ(G)$ is known in advance; at the same time, our upper bounds do not require $κ(G)$ to be given in advance.

cs.DS↗

arXiv2012

Advanced Bloom Filter Based Algorithms for Efficient Approximate Data De-Duplication in Streams

Applications involving telecommunication call data records, web pages, online transactions, medical records, stock markets, climate warning systems, etc., necessitate efficient management and processing of such massively exponential amount of data from diverse sources. De-duplication or Intelligent Compression in streaming scenarios for approximate identification and elimination of duplicates from such unbounded data stream is a greater challenge given the real-time nature of data arrival. Stable Bloom Filters (SBF) addresses this problem to a certain extent. . In this work, we present several novel algorithms for the problem of approximate detection of duplicates in data streams. We propose the Reservoir Sampling based Bloom Filter (RSBF) combining the working principle of reservoir sampling and Bloom Filters. We also present variants of the novel Biased Sampling based Bloom Filter (BSBF) based on biased sampling concepts. We also propose a randomized load balanced variant of the sampling Bloom Filter approach to efficiently tackle the duplicate detection. In this work, we thus provide a generic framework for de-duplication using Bloom Filters. Using detailed theoretical analysis we prove analytical bounds on the false positive rate, false negative rate and convergence rate of the proposed structures. We exhibit that our models clearly outperform the existing methods. We also demonstrate empirical analysis of the structures using real-world datasets (3 million records) and also with synthetic datasets (1 billion records) capturing various input distributions.

cs.IR↗