-
Geometry and Dress groups with non-symmetric cost functions
Authors:
Lukas Silvester Barth,
Parvaneh Joharinad,
Jürgen Jost,
Walter Wenzel
Abstract:
A metric relation by definition is symmetric. Since many data sets are non-symmetric, in this paper we develop a systematic theory of non-symmetric cost functions. Betweenness relations play an important role. We also introduce the notion of a Dress group in the non-symmetric setting and indicate a notion of curvature.
A metric relation by definition is symmetric. Since many data sets are non-symmetric, in this paper we develop a systematic theory of non-symmetric cost functions. Betweenness relations play an important role. We also introduce the notion of a Dress group in the non-symmetric setting and indicate a notion of curvature.
△ Less
Submitted 31 May, 2025;
originally announced June 2025.
-
Efficient compression of neural networks and datasets
Authors:
Lukas Silvester Barth,
Paulo von Petersenn
Abstract:
We compare, improve, and contribute methods that substantially decrease the number of parameters of neural networks while maintaining high test accuracy. When applying our methods to minimize description length, we obtain very effective data compression algorithms. In particular, we develop a probabilistic reformulation of $\ell_0$ regularized optimization for nonlinear models that does not requir…
▽ More
We compare, improve, and contribute methods that substantially decrease the number of parameters of neural networks while maintaining high test accuracy. When applying our methods to minimize description length, we obtain very effective data compression algorithms. In particular, we develop a probabilistic reformulation of $\ell_0$ regularized optimization for nonlinear models that does not require Monte-Carlo sampling and thus improves upon previous methods. We also improve upon methods involving smooth approximations to the $\ell_0$ norm, and investigate layerwise methods. We compare the methods on different architectures and datasets, including convolutional networks trained on image datasets and transformers trained on parts of Wikipedia. We also created a synthetic teacher-student setup to investigate compression in a controlled continuous setting. Finally, we conceptually relate compression algorithms to Solomonoff's theory of inductive inference and empirically verify the prediction that regularized models can exhibit more sample-efficient convergence.
△ Less
Submitted 23 May, 2025;
originally announced May 2025.
-
Merging Hazy Sets with m-Schemes: A Geometric Approach to Data Visualization
Authors:
Lukas Silvester Barth,
Hannaneh Fahimi,
Parvaneh Joharinad,
Jürgen Jost,
Janis Keck
Abstract:
Many machine learning algorithms try to visualize high dimensional metric data in 2D in such a way that the essential geometric and topological features of the data are highlighted. In this paper, we introduce a framework for aggregating dissimilarity functions that arise from locally adjusting a metric through density-aware normalization, as employed in the IsUMap method. We formalize these appro…
▽ More
Many machine learning algorithms try to visualize high dimensional metric data in 2D in such a way that the essential geometric and topological features of the data are highlighted. In this paper, we introduce a framework for aggregating dissimilarity functions that arise from locally adjusting a metric through density-aware normalization, as employed in the IsUMap method. We formalize these approaches as m-schemes, a class of methods closely related to t-norms and t-conorms in probabilistic metrics, as well as to composition laws in information theory. These m-schemes provide a flexible and theoretically grounded approach to refining distance-based embeddings.
△ Less
Submitted 3 March, 2025;
originally announced March 2025.
-
IsUMap: Manifold Learning and Data Visualization leveraging Vietoris-Rips filtrations
Authors:
Lukas Silvester Barth,
Fatemeh,
Fahimi,
Parvaneh Joharinad,
Jürgen Jost,
Janis Keck
Abstract:
This work introduces IsUMap, a novel manifold learning technique that enhances data representation by integrating aspects of UMAP and Isomap with Vietoris-Rips filtrations. We present a systematic and detailed construction of a metric representation for locally distorted metric spaces that captures complex data structures more accurately than the previous schemes. Our approach addresses limitation…
▽ More
This work introduces IsUMap, a novel manifold learning technique that enhances data representation by integrating aspects of UMAP and Isomap with Vietoris-Rips filtrations. We present a systematic and detailed construction of a metric representation for locally distorted metric spaces that captures complex data structures more accurately than the previous schemes. Our approach addresses limitations in existing methods by accommodating non-uniform data distributions and intricate local geometries. We validate its performance through extensive experiments on examples of various geometric objects and benchmark real-world datasets, demonstrating significant improvements in representation quality.
△ Less
Submitted 25 July, 2024;
originally announced July 2024.
-
A Topology-Shape-Metrics Framework for Ortho-Radial Graph Drawing
Authors:
Lukas Barth,
Benjamin Niedermann,
Ignaz Rutter,
Matthias Wolf
Abstract:
Orthogonal drawings, i.e., embeddings of graphs into grids, are a classic topic in Graph Drawing. Often the goal is to find a drawing that minimizes the number of bends on the edges. A key ingredient for bend minimization algorithms is the existence of an orthogonal representation that describes such drawings combinatorially by only listing the angles between the edges around each vertex and the d…
▽ More
Orthogonal drawings, i.e., embeddings of graphs into grids, are a classic topic in Graph Drawing. Often the goal is to find a drawing that minimizes the number of bends on the edges. A key ingredient for bend minimization algorithms is the existence of an orthogonal representation that describes such drawings combinatorially by only listing the angles between the edges around each vertex and the directions of bends on the edges, but neglecting any kind of geometric information such as vertex coordinates or edge lengths.
We generalize this idea to ortho-radial representations of ortho-radial drawings, which are embeddings into an ortho-radial grid, whose gridlines are concentric circles around the origin and straight-line spokes emanating from the origin but excluding the origin itself. Unlike the orthogonal case, there exist ortho-radial representations that do not admit a corresponding drawing, for example so-called strictly monotone cycles. An ortho-radial drawing is called valid if it does not contain a strictly monotone cycle. Our first result is that an ortho-radial representation admits a corresponding drawing if and only if it is valid. Previously such a characterization was only known for ortho-radial drawings of paths, cycles, and theta graphs, and in the special case of rectangular drawings of cubic graphs, where the contour of each face is required to be a rectangle. Further, we give a quadratic-time algorithm that tests for an ortho-radial representation whether it is valid, and we show how to draw a valid ortho-radial representation in the same running time.
Altogether, this reduces the problem of computing a minimum-bend ortho-radial drawing to the task of computing a valid ortho-radial representation with the minimum number of bends, and hence establishes an ortho-radial analogue of the topology-shape-metrics framework for planar orthogonal drawings by Tamassia.
△ Less
Submitted 10 June, 2021;
originally announced June 2021.
-
Zipping Segment Trees
Authors:
Lukas Barth,
Dorothea Wagner
Abstract:
Stabbing queries in sets of intervals are usually answered using segment trees. A dynamic variant of segment trees has been presented by van Kreveld and Overmars, which uses red-black trees to do rebalancing operations. This paper presents zipping segment trees - dynamic segment trees based on zip trees, which were recently introduced by Tarjan et al. To facilitate zipping segment trees, we show h…
▽ More
Stabbing queries in sets of intervals are usually answered using segment trees. A dynamic variant of segment trees has been presented by van Kreveld and Overmars, which uses red-black trees to do rebalancing operations. This paper presents zipping segment trees - dynamic segment trees based on zip trees, which were recently introduced by Tarjan et al. To facilitate zipping segment trees, we show how to uphold certain segment tree properties during the operations of a zip tree. We present an in-depth experimental evaluation and comparison of dynamic segment trees based on red-black trees, weight-balanced trees and several variants of the novel zipping segment trees. Our results indicate that zipping segment trees perform better than rotation-based alternatives.
△ Less
Submitted 7 April, 2020;
originally announced April 2020.
-
Engineering Top-Down Weight-Balanced Trees
Authors:
Lukas Barth,
Dorothea Wagner
Abstract:
Weight-balanced trees are a popular form of self-balancing binary search trees. Their popularity is due to desirable guarantees, for example regarding the required work to balance annotated trees.
While usual weight-balanced trees perform their balancing operations in a bottom-up fashion after a modification to the tree is completed, there exists a top-down variant which performs these balancing…
▽ More
Weight-balanced trees are a popular form of self-balancing binary search trees. Their popularity is due to desirable guarantees, for example regarding the required work to balance annotated trees.
While usual weight-balanced trees perform their balancing operations in a bottom-up fashion after a modification to the tree is completed, there exists a top-down variant which performs these balancing operations during descend. This variant has so far received only little attention. We provide an in-depth analysis and engineering of these top-down weight-balanced trees, demonstrating their superior performance. We also gaining insights into how the balancing parameters necessary for a weight-balanced tree should be chosen - with the surprising observation that it is often beneficial to choose parameters which are not feasible in the sense of the correctness proofs for the rebalancing algorithm.
△ Less
Submitted 28 October, 2019; v1 submitted 17 October, 2019;
originally announced October 2019.
-
Multilevel Planarity
Authors:
Lukas Barth,
Guido Brückner,
Paul Jungeblut,
Marcel Radermacher
Abstract:
In this paper, we introduce and study the multilevel-planarity testing problem, which is a generalization of upward planarity and level planarity. Let $G = (V, E)$ be a directed graph and let $\ell: V \to \mathcal P(\mathbb Z)$ be a function that assigns a finite set of integers to each vertex. A multilevel-planar drawing of $G$ is a planar drawing of $G$ such that the $y$-coordinate of each verte…
▽ More
In this paper, we introduce and study the multilevel-planarity testing problem, which is a generalization of upward planarity and level planarity. Let $G = (V, E)$ be a directed graph and let $\ell: V \to \mathcal P(\mathbb Z)$ be a function that assigns a finite set of integers to each vertex. A multilevel-planar drawing of $G$ is a planar drawing of $G$ such that the $y$-coordinate of each vertex $v \in V$ is $y(v) \in \ell(v)$, and each edge is drawn as a strictly $y$-monotone curve. We present linear-time algorithms for testing multilevel planarity of embedded graphs with a single source and of oriented cycles. Complementing these algorithmic results, we show that multilevel-planarity testing is NP-complete even in very restricted cases.
△ Less
Submitted 31 October, 2018;
originally announced October 2018.
-
Towards a Topology-Shape-Metrics Framework for Ortho-Radial Drawings
Authors:
Lukas Barth,
Benjamin Niedermann,
Ignaz Rutter,
Matthias Wolf
Abstract:
Ortho-Radial drawings are a generalization of orthogonal drawings to grids that are formed by concentric circles and straight-line spokes emanating from the circles' center. Such drawings have applications in schematic graph layouts, e.g., for metro maps and destination maps.
A plane graph is a planar graph with a fixed planar embedding. We give a combinatorial characterization of the plane grap…
▽ More
Ortho-Radial drawings are a generalization of orthogonal drawings to grids that are formed by concentric circles and straight-line spokes emanating from the circles' center. Such drawings have applications in schematic graph layouts, e.g., for metro maps and destination maps.
A plane graph is a planar graph with a fixed planar embedding. We give a combinatorial characterization of the plane graphs that admit a planar ortho-radial drawing without bends. Previously, such a characterization was only known for paths, cycles, and theta graphs, and in the special case of rectangular drawings for cubic graphs, where the contour of each face is required to be a rectangle.
The characterization is expressed in terms of an ortho-radial representation that, similar to Tamassia's orthogonal representations for orthogonal drawings describes such a drawing combinatorially in terms of angles around vertices and bends on the edges. In this sense our characterization can be seen as a first step towards generalizing the Topology-Shape-Metrics framework of Tamassia to ortho-radial drawings.
△ Less
Submitted 17 March, 2017;
originally announced March 2017.
-
Temporal Map Labeling: A New Unified Framework with Experiments
Authors:
Lukas Barth,
Benjamin Niedermann,
Martin Nöllenburg,
Darren Strash
Abstract:
The increased availability of interactive maps on the Internet and on personal mobile devices has created new challenges in computational cartography and, in particular, for label placement in maps. Operations like rotation, zoom, and translation dynamically change the map over time and make a consistent adaptation of the map labeling necessary.
In this paper, we consider map labeling for the ca…
▽ More
The increased availability of interactive maps on the Internet and on personal mobile devices has created new challenges in computational cartography and, in particular, for label placement in maps. Operations like rotation, zoom, and translation dynamically change the map over time and make a consistent adaptation of the map labeling necessary.
In this paper, we consider map labeling for the case that a map undergoes a sequence of operations over a specified time span. We unify and generalize several preceding models for dynamic map labeling into one versatile and flexible model. In contrast to previous research, we completely abstract from the particular operations (e.g., zoom, rotation, etc.) and express the labeling problem as a set of time intervals representing the labels' presences, activities, and conflicts. The model's strength is manifested in its simplicity and broad range of applications. In particular, it supports label selection both for map features with fixed position as well as for moving entities (e.g., for tracking vehicles in logistics or air traffic control).
Through extensive experiments on OpenStreetMap data, we evaluate our model using algorithms of varying complexity as a case study for navigation systems. Our experiments show that even simple (and thus, fast) algorithms achieve near-optimal solutions in our model with respect to an intuitive objective function.
△ Less
Submitted 20 September, 2016;
originally announced September 2016.
-
On the Readability of Boundary Labeling
Authors:
Lukas Barth,
Andreas Gemsa,
Benjamin Niedermann,
Martin Nöllenburg
Abstract:
Boundary labeling deals with annotating features in images such that labels are placed outside of the image and are connected by curves (so-called leaders) to the corresponding features. While boundary labeling has been extensively investigated from an algorithmic perspective, the research on its readability has been neglected. In this paper we present the first formal user study on the readabilit…
▽ More
Boundary labeling deals with annotating features in images such that labels are placed outside of the image and are connected by curves (so-called leaders) to the corresponding features. While boundary labeling has been extensively investigated from an algorithmic perspective, the research on its readability has been neglected. In this paper we present the first formal user study on the readability of boundary labeling. We consider the four most studied leader types with respect to their performance, i.e., whether and how fast a viewer can assign a feature to its label and vice versa. We give a detailed analysis of the results regarding the readability of the four models and discuss their aesthetic qualities based on the users' preference judgments and interviews.
△ Less
Submitted 1 September, 2015;
originally announced September 2015.
-
Semantic Word Cloud Representations: Hardness and Approximation Algorithms
Authors:
Lukas Barth,
Sara Irina Fabrikant,
Stephen Kobourov,
Anna Lubiw,
Martin Nöllenburg,
Yoshio Okamoto,
Sergey Pupyrev,
Claudio Squarcella,
Torsten Ueckerdt,
Alexander Wolff
Abstract:
We study a geometric representation problem, where we are given a set $\cal R$ of axis-aligned rectangles with fixed dimensions and a graph with vertex set $\cal R$. The task is to place the rectangles without overlap such that two rectangles touch if and only if the graph contains an edge between them. We call this problem Contact Representation of Word Networks (CROWN). It formalizes the geometr…
▽ More
We study a geometric representation problem, where we are given a set $\cal R$ of axis-aligned rectangles with fixed dimensions and a graph with vertex set $\cal R$. The task is to place the rectangles without overlap such that two rectangles touch if and only if the graph contains an edge between them. We call this problem Contact Representation of Word Networks (CROWN). It formalizes the geometric problem behind drawing word clouds in which semantically related words are close to each other. Here, we represent words by rectangles and semantic relationships by edges. We show that CROWN is strongly NP-hard even restricted trees and weakly NP-hard if restricted stars. We consider the optimization problem Max-CROWN where each adjacency induces a certain profit and the task is to maximize the sum of the profits. For this problem, we present constant-factor approximations for several graph classes, namely stars, trees, planar graphs, and graphs of bounded degree. Finally, we evaluate the algorithms experimentally and show that our best method improves upon the best existing heuristic by 45%.
△ Less
Submitted 19 November, 2013;
originally announced November 2013.
-
On Semantic Word Cloud Representation
Authors:
Lukas Barth,
Stephen Kobourov,
Sergey Pupyrev,
Torsten Ueckerdt
Abstract:
We study the problem of computing semantic-preserving word clouds in which semantically related words are close to each other. While several heuristic approaches have been described in the literature, we formalize the underlying geometric algorithm problem: Word Rectangle Adjacency Contact (WRAC). In this model each word is associated with rectangle with fixed dimensions, and the goal is to repres…
▽ More
We study the problem of computing semantic-preserving word clouds in which semantically related words are close to each other. While several heuristic approaches have been described in the literature, we formalize the underlying geometric algorithm problem: Word Rectangle Adjacency Contact (WRAC). In this model each word is associated with rectangle with fixed dimensions, and the goal is to represent semantically related words by ensuring that the two corresponding rectangles touch. We design and analyze efficient polynomial-time algorithms for some variants of the WRAC problem, show that several general variants are NP-hard, and describe a number of approximation algorithms. Finally, we experimentally demonstrate that our theoretically-sound algorithms outperform the early heuristics.
△ Less
Submitted 23 April, 2013;
originally announced April 2013.
