Showing 1–49 of 49 results for author: Schubert, E

Explicit Formulae to Interchangeably use Hyperplanes and Hyperballs using Inversive Geometry

Authors: Erik Thordsen, Erich Schubert

Abstract: Many algorithms require discriminative boundaries, such as separating hyperplanes or hyperballs, or are specifically designed to work on spherical data. By applying inversive geometry, we show that the two discriminative boundaries can be used interchangeably, and that general Euclidean data can be transformed into spherical data, whenever a change in point distances is acceptable. We provide expl… ▽ More Many algorithms require discriminative boundaries, such as separating hyperplanes or hyperballs, or are specifically designed to work on spherical data. By applying inversive geometry, we show that the two discriminative boundaries can be used interchangeably, and that general Euclidean data can be transformed into spherical data, whenever a change in point distances is acceptable. We provide explicit formulae to embed general Euclidean data into spherical data and to unembed it back. We further show a duality between hyperspherical caps, i.e., the volume created by a separating hyperplane on spherical data, and hyperballs and provide explicit formulae to map between the two. We further provide equations to translate inner products and Euclidean distances between the two spaces, to avoid explicit embedding and unembedding. We also provide a method to enforce projections of the general Euclidean space onto hemi-hyperspheres and propose an intrinsic dimensionality based method to obtain "all-purpose" parameters. To show the usefulness of the cap-ball-duality, we discuss example applications in machine learning and vector similarity search. △ Less

Submitted 28 May, 2024; originally announced May 2024.

Comments: to be submitted to TMLR (submission pending)

Controlling the broadband enhanced light chirality with L-shaped dielectric metamaterials

Authors: Ufuk Kilic, Matthew Hilfiker, Shawn Wimer, Alexander Ruder, Eva Schubert, Mathias Schubert, Christos Argyropoulos

Abstract: The inherently weak chiroptical responses of natural materials limit their usage for controlling and enhancing chiral light-matter interactions. Recently, several nanostructures with subwavelength scale dimensions were demonstrated, mainly due to the advent of nanofabrication technologies, as a potential alternative to efficiently enhance chirality. However, the intrinsic lossy nature of metals an… ▽ More The inherently weak chiroptical responses of natural materials limit their usage for controlling and enhancing chiral light-matter interactions. Recently, several nanostructures with subwavelength scale dimensions were demonstrated, mainly due to the advent of nanofabrication technologies, as a potential alternative to efficiently enhance chirality. However, the intrinsic lossy nature of metals and inherent narrowband response of dielectric planar thin films or metasurface structures pose severe limitations toward the practical realization of broadband and tailorable chiral systems. Here, we tackle these problems by designing all-dielectric silicon-based L-shaped optical metamaterials based on tilted nanopillars that exhibit broadband and enhanced chiroptical response in transmission operation. We use an emerging bottom-up fabrication approach, named glancing angle deposition, to assemble these dielectric metamaterials on a wafer scale. The reported strong chirality and optical anisotropic properties are controllable in terms of both amplitude and operating frequency by simply varying the shape and dimensions of the nanopillars. The presented nanostructures can be used in a plethora of emerging nanophotonic applications, such as chiral sensors, polarization filters, and spin-locked nanowaveguides. △ Less

Submitted 27 March, 2024; originally announced March 2024.

Journal ref: U. Kilic, M. Hilfiker, S. Wimer, A. Ruder, E. Schubert, M. Schubert, and C. Argyropoulos, Controlling the broadband enhanced light chirality with L-shaped dielectric metamaterials, Nature Communications, vol. 15, p. 3757, 2024

Nanocolumnar Material Platforms:Universal structural parameters revealed from optical anisotropy

Authors: Ufuk Kilic, Yousra Traouli, Matthew Hilfiker, Khalil Bryant, Stefan Schoeche, Rene Feder, Christos Argyropoulos, Eva Schubert, Mathias Schubert

Abstract: Nanostructures represent a frontier where meticulous attention to the control and assessment of structural dimensions becomes a linchpin for their seamless integration into diverse technological applications. By using integrative and comprehensive methodical series of studies, we investigate the evolution of the depolarization factors in the anisotropic Bruggeman effective medium approximation, th… ▽ More Nanostructures represent a frontier where meticulous attention to the control and assessment of structural dimensions becomes a linchpin for their seamless integration into diverse technological applications. By using integrative and comprehensive methodical series of studies, we investigate the evolution of the depolarization factors in the anisotropic Bruggeman effective medium approximation, that are extremely sensitive to the changes in critical dimensions of the nanostructure platforms. To this end, we fabricate spatially coherent highly-ordered slanted nanocolumns from zirconia, silicon, titanium, and permalloy on silicon substrates with varying column lengths using glancing angle deposition. In tandem, broad-spectral range Mueller matrix spectroscopic ellipsometry data, spanning from the near-infrared to the vacuum ultraviolet (0.72 eV to 6.5 eV), is analyzed with a best-match model approach based on the anisotropic Bruggeman effective medium theory. We thereby extracted the anisotropic optical properties including complex dielectric function, birefringence, and dichroism. Most notably, our research unveils a universal, material-independent inverse relationship between depolarization factors and column length. We envision that the presented universal relationship will permit accurate prediction of optical properties of nanocolumnar thin films improving their integration and optimization for optoelectronic and photonic device applications. △ Less

Submitted 1 December, 2023; originally announced December 2023.

Comments: 20 pages, 10 figures

Medoid Silhouette clustering with automatic cluster number selection

Authors: Lars Lenssen, Erich Schubert

Abstract: The evaluation of clustering results is difficult, highly dependent on the evaluated data set and the perspective of the beholder. There are many different clustering quality measures, which try to provide a general measure to validate clustering results. A very popular measure is the Silhouette. We discuss the efficient medoid-based variant of the Silhouette, perform a theoretical analysis of its… ▽ More The evaluation of clustering results is difficult, highly dependent on the evaluated data set and the perspective of the beholder. There are many different clustering quality measures, which try to provide a general measure to validate clustering results. A very popular measure is the Silhouette. We discuss the efficient medoid-based variant of the Silhouette, perform a theoretical analysis of its properties, provide two fast versions for the direct optimization, and discuss the use to choose the optimal number of clusters. We combine ideas from the original Silhouette with the well-known PAM algorithm and its latest improvements FasterPAM. One of the versions guarantees equal results to the original variant and provides a run speedup of $O(k^2)$. In experiments on real data with 30000 samples and $k$=100, we observed a 10464$\times$ speedup compared to the original PAMMEDSIL algorithm. Additionally, we provide a variant to choose the optimal number of clusters directly. △ Less

Submitted 7 September, 2023; originally announced September 2023.

Comments: arXiv admin note: substantial text overlap with arXiv:2209.12553

Sparse Partitioning Around Medoids

Authors: Lars Lenssen, Erich Schubert

Abstract: Partitioning Around Medoids (PAM, k-Medoids) is a popular clustering technique to use with arbitrary distance functions or similarities, where each cluster is represented by its most central object, called the medoid or the discrete median. In operations research, this family of problems is also known as facility location problem (FLP). FastPAM recently introduced a speedup for large k to make it… ▽ More Partitioning Around Medoids (PAM, k-Medoids) is a popular clustering technique to use with arbitrary distance functions or similarities, where each cluster is represented by its most central object, called the medoid or the discrete median. In operations research, this family of problems is also known as facility location problem (FLP). FastPAM recently introduced a speedup for large k to make it applicable for larger problems, but the method still has a runtime quadratic in N. In this chapter, we discuss a sparse and asymmetric variant of this problem, to be used for example on graph data such as road networks. By exploiting sparsity, we can avoid the quadratic runtime and memory requirements, and make this method scalable to even larger problems, as long as we are able to build a small enough graph of sufficient connectivity to perform local optimization. Furthermore, we consider asymmetric cases, where the set of medoids is not identical to the set of points to be covered (or in the interpretation of facility location, where the possible facility locations are not identical to the consumer locations). Because of sparsity, it may be impossible to cover all points with just k medoids for too small k, which would render the problem unsolvable, and this breaks common heuristics for finding a good starting condition. We, hence, consider determining k as a part of the optimization problem and propose to first construct a greedy initial solution with a larger k, then to optimize the problem by alternating between PAM-style "swap" operations where the result is improved by replacing medoids with better alternatives and "remove" operations to reduce the number of k until neither allows further improving the result quality. We demonstrate the usefulness of this method on a problem from electrical engineering, with the input graph derived from cartographic data. △ Less

Submitted 5 September, 2023; originally announced September 2023.

Data Aggregation for Hierarchical Clustering

Authors: Erich Schubert, Andreas Lang

Abstract: Hierarchical Agglomerative Clustering (HAC) is likely the earliest and most flexible clustering method, because it can be used with many distances, similarities, and various linkage strategies. It is often used when the number of clusters the data set forms is unknown and some sort of hierarchy in the data is plausible. Most algorithms for HAC operate on a full distance matrix, and therefore requi… ▽ More Hierarchical Agglomerative Clustering (HAC) is likely the earliest and most flexible clustering method, because it can be used with many distances, similarities, and various linkage strategies. It is often used when the number of clusters the data set forms is unknown and some sort of hierarchy in the data is plausible. Most algorithms for HAC operate on a full distance matrix, and therefore require quadratic memory. The standard algorithm also has cubic runtime to produce a full hierarchy. Both memory and runtime are especially problematic in the context of embedded or otherwise very resource-constrained systems. In this section, we present how data aggregation with BETULA, a numerically stable version of the well known BIRCH data aggregation algorithm, can be used to make HAC viable on systems with constrained resources with only small losses on clustering quality, and hence allow exploratory data analysis of very large data sets. △ Less

Submitted 5 September, 2023; originally announced September 2023.

LOSDD: Leave-Out Support Vector Data Description for Outlier Detection

Authors: Daniel Boiar, Thomas Liebig, Erich Schubert

Abstract: Support Vector Machines have been successfully used for one-class classification (OCSVM, SVDD) when trained on clean data, but they work much worse on dirty data: outliers present in the training data tend to become support vectors, and are hence considered "normal". In this article, we improve the effectiveness to detect outliers in dirty training data with a leave-out strategy: by temporarily om… ▽ More Support Vector Machines have been successfully used for one-class classification (OCSVM, SVDD) when trained on clean data, but they work much worse on dirty data: outliers present in the training data tend to become support vectors, and are hence considered "normal". In this article, we improve the effectiveness to detect outliers in dirty training data with a leave-out strategy: by temporarily omitting one candidate at a time, this point can be judged using the remaining data only. We show that this is more effective at scoring the outlierness of points than using the slack term of existing SVM-based approaches. Identified outliers can then be removed from the data, such that outliers hidden by other outliers can be identified, to reduce the problem of masking. Naively, this approach would require training N individual SVMs (and training $O(N^2)$ SVMs when iteratively removing the worst outliers one at a time), which is prohibitively expensive. We will discuss that only support vectors need to be considered in each step and that by reusing SVM parameters and weights, this incremental retraining can be accelerated substantially. By removing candidates in batches, we can further improve the processing time, although it obviously remains more costly than training a single SVM. △ Less

Submitted 27 December, 2022; originally announced December 2022.

Stop using the elbow criterion for k-means and how to choose the number of clusters instead

Authors: Erich Schubert

Abstract: A major challenge when using k-means clustering often is how to choose the parameter k, the number of clusters. In this letter, we want to point out that it is very easy to draw poor conclusions from a common heuristic, the "elbow method". Better alternatives have been known in literature for a long time, and we want to draw attention to some of these easy to use options, that often perform better… ▽ More A major challenge when using k-means clustering often is how to choose the parameter k, the number of clusters. In this letter, we want to point out that it is very easy to draw poor conclusions from a common heuristic, the "elbow method". Better alternatives have been known in literature for a long time, and we want to draw attention to some of these easy to use options, that often perform better. This letter is a call to stop using the elbow method altogether, because it severely lacks theoretic support, and we want to encourage educators to discuss the problems of the method -- if introducing it in class at all -- and teach alternatives instead, while researchers and reviewers should reject conclusions drawn from the elbow method. △ Less

Submitted 23 December, 2022; originally announced December 2022.

Algebra of N-event synchronization

Authors: Ernesto Gomez, Keith E. Schubert, Khalil Dajani

Abstract: We have previously defined synchronization (Gomez, E. and K. Schubert 2011) as a relation between the times at which a pair of events can happen, and introduced an algebra that covers all possible relations for such pairs. In this work we introduce the synchronization matrix, to make it easier to calculate the properties and results of $N$ event synchronizations, such as are commonly encountered i… ▽ More We have previously defined synchronization (Gomez, E. and K. Schubert 2011) as a relation between the times at which a pair of events can happen, and introduced an algebra that covers all possible relations for such pairs. In this work we introduce the synchronization matrix, to make it easier to calculate the properties and results of $N$ event synchronizations, such as are commonly encountered in parallel execution of multiple processes. The synchronization matrix leads to the definition of N-event synchronization algebras as specific extensions to the original algebra. We derive general properties of such synchronization, and we are able to analyze effects of synchronization on the phase space of parallel execution introduced in (Gomez E Kai R, Schubert KE 2017) △ Less

Submitted 1 November, 2022; originally announced November 2022.

Comments: 9 pages, 2 figures

ACM Class: B.4.3; D.3.1; D.3.2; D.3.3

Clustering by Direct Optimization of the Medoid Silhouette

Authors: Lars Lenssen, Erich Schubert

Abstract: The evaluation of clustering results is difficult, highly dependent on the evaluated data set and the perspective of the beholder. There are many different clustering quality measures, which try to provide a general measure to validate clustering results. A very popular measure is the Silhouette. We discuss the efficient medoid-based variant of the Silhouette, perform a theoretical analysis of its… ▽ More The evaluation of clustering results is difficult, highly dependent on the evaluated data set and the perspective of the beholder. There are many different clustering quality measures, which try to provide a general measure to validate clustering results. A very popular measure is the Silhouette. We discuss the efficient medoid-based variant of the Silhouette, perform a theoretical analysis of its properties, and provide two fast versions for the direct optimization. We combine ideas from the original Silhouette with the well-known PAM algorithm and its latest improvements FasterPAM. One of the versions guarantees equal results to the original variant and provides a run speedup of $O(k^2)$. In experiments on real data with 30000 samples and $k$=100, we observed a 10464$\times$ speedup compared to the original PAMMEDSIL algorithm. △ Less

Submitted 26 September, 2022; originally announced September 2022.

On Projections to Linear Subspaces

Authors: Erik Thordsen, Erich Schubert

Abstract: The merit of projecting data onto linear subspaces is well known from, e.g., dimension reduction. One key aspect of subspace projections, the maximum preservation of variance (principal component analysis), has been thoroughly researched and the effect of random linear projections on measures such as intrinsic dimensionality still is an ongoing effort. In this paper, we investigate the less explor… ▽ More The merit of projecting data onto linear subspaces is well known from, e.g., dimension reduction. One key aspect of subspace projections, the maximum preservation of variance (principal component analysis), has been thoroughly researched and the effect of random linear projections on measures such as intrinsic dimensionality still is an ongoing effort. In this paper, we investigate the less explored depths of linear projections onto explicit subspaces of varying dimensionality and the expectations of variance that ensue. The result is a new family of bounds for Euclidean distances and inner products. We showcase the quality of these bounds as well as investigate the intimate relation to intrinsic dimensionality estimation. △ Less

Submitted 26 September, 2022; originally announced September 2022.

Dilated POCS: Minimax Convex Optimization

Authors: Albert R. Yu, Robert J. Marks II, Keith E. Schubert, Charles Baylis, Austin Egbert, Adam Goad, Sam Haug

Abstract: Alternating projection onto convex sets (POCS) provides an iterative procedure to find a signal that satisfies two or more convex constraints when the sets intersect. For nonintersecting constraints, the method of simultaneous projections produces a minimum mean square error (MMSE) solution. In certain cases, a minimax solution is more desirable. Generating a minimax solution is possible using dil… ▽ More Alternating projection onto convex sets (POCS) provides an iterative procedure to find a signal that satisfies two or more convex constraints when the sets intersect. For nonintersecting constraints, the method of simultaneous projections produces a minimum mean square error (MMSE) solution. In certain cases, a minimax solution is more desirable. Generating a minimax solution is possible using dilated POCS. The minimax solution uses morphological dilation of nonintersecting signal convex constraints. The sets are progressively dilated to the point where there is intersection at a minimax solution. Examples are given contrasting the MMSE and minimax solutions in problems of tomographic reconstruction of images. Dilated POCS adds a new imaging modality for image synthesis. Lastly, morphological erosion of signal sets is suggested as a method to shrink the overlap when sets intersect at more than one point. △ Less

Submitted 27 January, 2023; v1 submitted 9 June, 2022; originally announced June 2022.

Comments: 8 pages, 12 figures

EmbAssi: Embedding Assignment Costs for Similarity Search in Large Graph Databases

Authors: Franka Bause, Erich Schubert, Nils M. Kriege

Abstract: The graph edit distance is an intuitive measure to quantify the dissimilarity of graphs, but its computation is NP-hard and challenging in practice. We introduce methods for answering nearest neighbor and range queries regarding this distance efficiently for large databases with up to millions of graphs. We build on the filter-verification paradigm, where lower and upper bounds are used to reduce… ▽ More The graph edit distance is an intuitive measure to quantify the dissimilarity of graphs, but its computation is NP-hard and challenging in practice. We introduce methods for answering nearest neighbor and range queries regarding this distance efficiently for large databases with up to millions of graphs. We build on the filter-verification paradigm, where lower and upper bounds are used to reduce the number of exact computations of the graph edit distance. Highly effective bounds for this involve solving a linear assignment problem for each graph in the database, which is prohibitive in massive datasets. Index-based approaches typically provide only weak bounds leading to high computational costs verification. In this work, we derive novel lower bounds for efficient filtering from restricted assignment problems, where the cost function is a tree metric. This special case allows embedding the costs of optimal assignments isometrically into $\ell_1$ space, rendering efficient indexing possible. We propose several lower bounds of the graph edit distance obtained from tree metrics reflecting the edit costs, which are combined for effective filtering. Our method termed EmbAssi can be integrated into existing filter-verification pipelines as a fast and effective pre-filtering step. Empirically we show that for many real-world graphs our lower bounds are already close to the exact graph edit distance, while our index construction and search scales to very large databases. △ Less

Submitted 19 July, 2022; v1 submitted 15 November, 2021; originally announced November 2021.

Comments: Data Min Knowl Disc (2022)

Metric Indexing for Graph Similarity Search

Authors: Franka Bause, David B. Blumenthal, Erich Schubert, Nils M. Kriege

Abstract: Finding the graphs that are most similar to a query graph in a large database is a common task with various applications. A widely-used similarity measure is the graph edit distance, which provides an intuitive notion of similarity and naturally supports graphs with vertex and edge attributes. Since its computation is NP-hard, techniques for accelerating similarity search have been studied extensi… ▽ More Finding the graphs that are most similar to a query graph in a large database is a common task with various applications. A widely-used similarity measure is the graph edit distance, which provides an intuitive notion of similarity and naturally supports graphs with vertex and edge attributes. Since its computation is NP-hard, techniques for accelerating similarity search have been studied extensively. However, index-based approaches for this are almost exclusively designed for graphs with categorical vertex and edge labels and uniform edit costs. We propose a filter-verification framework for similarity search, which supports non-uniform edit costs for graphs with arbitrary attributes. We employ an expensive lower bound obtained by solving an optimal assignment problem. This filter distance satisfies the triangle inequality, making it suitable for acceleration by metric indexing. In subsequent stages, assignment-based upper bounds are used to avoid further exact distance computations. Our extensive experimental evaluation shows that a significant runtime advantage over both a linear scan and state-of-the-art methods is achieved. △ Less

Submitted 4 October, 2021; originally announced October 2021.

Comments: SISAP 2021

Developments in Mathematical Algorithms and Computational Tools for Proton CT and Particle Therapy Treatment Planning

Authors: Yair Censor, Keith E. Schubert, Reinhard W. Schulte

Abstract: We summarize recent results and ongoing activities in mathematical algorithms and computer science methods related to proton computed tomography (pCT) and intensity-modulated particle therapy (IMPT) treatment planning. Proton therapy necessitates a high level of delivery accuracy to exploit the selective targeting imparted by the Bragg peak. For this purpose, pCT utilizes the proton beam itself to… ▽ More We summarize recent results and ongoing activities in mathematical algorithms and computer science methods related to proton computed tomography (pCT) and intensity-modulated particle therapy (IMPT) treatment planning. Proton therapy necessitates a high level of delivery accuracy to exploit the selective targeting imparted by the Bragg peak. For this purpose, pCT utilizes the proton beam itself to create images. The technique works by sending a low-intensity beam of protons through the patient and measuring the position, direction, and energy loss of each exiting proton. The pCT technique allows reconstruction of the volumetric distribution of the relative stop** power (RSP) of the patient tissues for use in treatment planning and pre-treatment range verification. We have investigated new ways to make the reconstruction both efficient and accurate. Better accuracy of RSP also enables more robust inverse approaches to IMPT. For IMPT, we developed a framework for performing intensity-modulation of the proton pencil beams. We expect that these developments will lead to additional project work in the years to come, which requires a regular exchange between experts in the fields of mathematics, computer science, and medical physics. We have initiated such an exchange by organizing annual workshops on pCT and IMPT algorithm and technology developments. This report is, admittedly, tilted toward our interdisciplinary work and methods. We offer a comprehensive overview of results, problems, and challenges in pCT and IMPT with the aim of making other scientists wanting to tackle such issues and to strengthen their interdisciplinary collaboration by bringing together cutting-edge know-how from medicine, computer science, physics, and mathematics to bear on medical physics problems at hand. △ Less

Submitted 21 August, 2021; originally announced August 2021.

Comments: 13 pages. Accepted for publication in IEEE Transactions on Radiation and Plasma Medical Sciences. August 16, 2021

MESS: Manifold Embedding Motivated Super Sampling

Authors: Erik Thordsen, Erich Schubert

Abstract: Many approaches in the field of machine learning and data analysis rely on the assumption that the observed data lies on lower-dimensional manifolds. This assumption has been verified empirically for many real data sets. To make use of this manifold assumption one generally requires the manifold to be locally sampled to a certain density such that features of the manifold can be observed. However,… ▽ More Many approaches in the field of machine learning and data analysis rely on the assumption that the observed data lies on lower-dimensional manifolds. This assumption has been verified empirically for many real data sets. To make use of this manifold assumption one generally requires the manifold to be locally sampled to a certain density such that features of the manifold can be observed. However, for increasing intrinsic dimensionality of a data set the required data density introduces the need for very large data sets, resulting in one of the many faces of the curse of dimensionality. To combat the increased requirement for local data density we propose a framework to generate virtual data points that faithful to an approximate embedding function underlying the manifold observable in the data. △ Less

Submitted 14 July, 2021; originally announced July 2021.

Accelerating Spherical k-Means

Authors: Erich Schubert, Andreas Lang, Gloria Feher

Abstract: Spherical k-means is a widely used clustering algorithm for sparse and high-dimensional data such as document vectors. While several improvements and accelerations have been introduced for the original k-means algorithm, not all easily translate to the spherical variant: Many acceleration techniques, such as the algorithms of Elkan and Hamerly, rely on the triangle inequality of Euclidean distance… ▽ More Spherical k-means is a widely used clustering algorithm for sparse and high-dimensional data such as document vectors. While several improvements and accelerations have been introduced for the original k-means algorithm, not all easily translate to the spherical variant: Many acceleration techniques, such as the algorithms of Elkan and Hamerly, rely on the triangle inequality of Euclidean distances. However, spherical k-means uses Cosine similarities instead of distances for computational efficiency. In this paper, we incorporate the Elkan and Hamerly accelerations to the spherical k-means algorithm working directly with the Cosines instead of Euclidean distances to obtain a substantial speedup and evaluate these spherical accelerations on real data. △ Less

Submitted 8 July, 2021; originally announced July 2021.

A Triangle Inequality for Cosine Similarity

Authors: Erich Schubert

Abstract: Similarity search is a fundamental problem for many data analysis techniques. Many efficient search techniques rely on the triangle inequality of metrics, which allows pruning parts of the search space based on transitive bounds on distances. Recently, Cosine similarity has become a popular alternative choice to the standard Euclidean metric, in particular in the context of textual data and neural… ▽ More Similarity search is a fundamental problem for many data analysis techniques. Many efficient search techniques rely on the triangle inequality of metrics, which allows pruning parts of the search space based on transitive bounds on distances. Recently, Cosine similarity has become a popular alternative choice to the standard Euclidean metric, in particular in the context of textual data and neural network embeddings. Unfortunately, Cosine similarity is not metric and does not satisfy the standard triangle inequality. Instead, many search techniques for Cosine rely on approximation techniques such as locality sensitive hashing. In this paper, we derive a triangle inequality for Cosine similarity that is suitable for efficient similarity search with many standard search structures (such as the VP-tree, Cover-tree, and M-tree); show that this bound is tight and discuss fast approximations for it. We hope that this spurs new research on accelerating exact similarity search for cosine similarity, and possible other similarity measures beyond the existing work for distance metrics. △ Less

Submitted 8 July, 2021; originally announced July 2021.

Broadband enhanced chirality with tunable response in hybrid plasmonic helical metamaterials

Authors: Ufuk Kilic, Matthew Hilfiker, Alexander Ruder, Rene Feder, Eva Schubert, Mathias Schubert, Christos Argyropoulos

Abstract: Designing broadband enhanced chirality is of strong interest to the emerging fields of chiral chemistry and sensing, or to control the spin orbital momentum of photons in recently introduced nanophotonic chiral quantum and classical optical applications. However, chiral light-matter interactions have an extremely weak nature, are difficult to be controlled and enhanced, and cannot be made tunable… ▽ More Designing broadband enhanced chirality is of strong interest to the emerging fields of chiral chemistry and sensing, or to control the spin orbital momentum of photons in recently introduced nanophotonic chiral quantum and classical optical applications. However, chiral light-matter interactions have an extremely weak nature, are difficult to be controlled and enhanced, and cannot be made tunable or broadband. In addition, planar ultrathin nanophotonic structures to achieve strong, broadband, and tunable chirality at the technologically important visible to ultraviolet spectrum still remain elusive. Here, we tackle these important problems by experimentally demonstrating and theoretically verifying spectrally tunable, extremely large, and broadband chiroptical response by nanohelical metamaterials. The reported new designs of all-dielectric and dielectric-metallic (hybrid) plasmonic metamaterials permit the largest and broadest ever measured chiral Kuhn dissymmetry factor achieved by a large-scale nanophotonic structure. In addition, the strong circular dichroism of the presented bottom-up fabricated optical metamaterials can be tuned by varying their dimensions and proportions between their dielectric and plasmonic helical subsections. The currently demonstrated ultrathin optical metamaterials are expected to provide a substantial boost to the develo** field of chiroptics leading to significantly enhanced and broadband chiral light-matter interactions at the nanoscale. △ Less

Submitted 28 January, 2021; v1 submitted 27 January, 2021; originally announced January 2021.

Journal ref: Adv. Funct. Mat., vol. 31, No. 20, p. 2010329, 2021

Fast and Eager k-Medoids Clustering: O(k) Runtime Improvement of the PAM, CLARA, and CLARANS Algorithms

Authors: Erich Schubert, Peter J. Rousseeuw

Abstract: Clustering non-Euclidean data is difficult, and one of the most used algorithms besides hierarchical clustering is the popular algorithm Partitioning Around Medoids (PAM), also simply referred to as k-medoids clustering. In Euclidean geometry the mean-as used in k-means-is a good estimator for the cluster center, but this does not exist for arbitrary dissimilarities. PAM uses the medoid instead, t… ▽ More Clustering non-Euclidean data is difficult, and one of the most used algorithms besides hierarchical clustering is the popular algorithm Partitioning Around Medoids (PAM), also simply referred to as k-medoids clustering. In Euclidean geometry the mean-as used in k-means-is a good estimator for the cluster center, but this does not exist for arbitrary dissimilarities. PAM uses the medoid instead, the object with the smallest dissimilarity to all others in the cluster. This notion of centrality can be used with any (dis-)similarity, and thus is of high relevance to many domains and applications. A key issue with PAM is its high run time cost. We propose modifications to the PAM algorithm that achieve an O(k)-fold speedup in the second ("SWAP") phase of the algorithm, but will still find the same results as the original PAM algorithm. If we relax the choice of swaps performed (while retaining comparable quality), we can further accelerate the algorithm by eagerly performing additional swaps in each iteration. With the substantially faster SWAP, we can now explore faster initialization strategies, because (i) the classic ("BUILD") initialization now becomes the bottleneck, and (ii) our swap is fast enough to compensate for worse starting conditions. We also show how the CLARA and CLARANS algorithms benefit from the proposed modifications. While we do not study the parallelization of our approach in this work, it can easily be combined with earlier approaches to use PAM and CLARA on big data (some of which use PAM as a subroutine, hence can immediately benefit from these improvements), where the performance with high k becomes increasingly important. In experiments on real data with k=100,200, we observed a 458x respectively 1191x speedup compared to the original PAM SWAP algorithm, making PAM applicable to larger data sets, and in particular to higher k. △ Less

Submitted 1 June, 2021; v1 submitted 12 August, 2020; originally announced August 2020.

Journal ref: Information Systems 2021, 101804

BETULA: Numerically Stable CF-Trees for BIRCH Clustering

Authors: Andreas Lang, Erich Schubert

Abstract: BIRCH clustering is a widely known approach for clustering, that has influenced much subsequent research and commercial products. The key contribution of BIRCH is the Clustering Feature tree (CF-Tree), which is a compressed representation of the input data. As new data arrives, the tree is eventually rebuilt to increase the compression. Afterward, the leaves of the tree are used for clustering. Be… ▽ More BIRCH clustering is a widely known approach for clustering, that has influenced much subsequent research and commercial products. The key contribution of BIRCH is the Clustering Feature tree (CF-Tree), which is a compressed representation of the input data. As new data arrives, the tree is eventually rebuilt to increase the compression. Afterward, the leaves of the tree are used for clustering. Because of the data compression, this method is very scalable. The idea has been adopted for example for k-means, data stream, and density-based clustering. Clustering features used by BIRCH are simple summary statistics that can easily be updated with new data: the number of points, the linear sums, and the sum of squared values. Unfortunately, how the sum of squares is then used in BIRCH is prone to catastrophic cancellation. We introduce a replacement cluster feature that does not have this numeric problem, that is not much more expensive to maintain, and which makes many computations simpler and hence more efficient. These cluster features can also easily be used in other work derived from BIRCH, such as algorithms for streaming data. In the experiments, we demonstrate the numerical problem and compare the performance of the original algorithm compared to the improved cluster features. △ Less

Submitted 23 June, 2020; originally announced June 2020.

ABID: Angle Based Intrinsic Dimensionality

Authors: Erik Thordsen, Erich Schubert

Abstract: The intrinsic dimensionality refers to the ``true'' dimensionality of the data, as opposed to the dimensionality of the data representation. For example, when attributes are highly correlated, the intrinsic dimensionality can be much lower than the number of variables. Local intrinsic dimensionality refers to the observation that this property can vary for different parts of the data set; and intr… ▽ More The intrinsic dimensionality refers to the ``true'' dimensionality of the data, as opposed to the dimensionality of the data representation. For example, when attributes are highly correlated, the intrinsic dimensionality can be much lower than the number of variables. Local intrinsic dimensionality refers to the observation that this property can vary for different parts of the data set; and intrinsic dimensionality can serve as a proxy for the local difficulty of the data set. Most popular methods for estimating the local intrinsic dimensionality are based on distances, and the rate at which the distances to the nearest neighbors increase, a concept known as ``expansion dimension''. In this paper we introduce an orthogonal concept, which does not use any distances: we use the distribution of angles between neighbor points. We derive the theoretical distribution of angles and use this to construct an estimator for intrinsic dimensionality. Experimentally, we verify that this measure behaves similarly, but complementarily, to existing measures of intrinsic dimensionality. By introducing a new idea of intrinsic dimensionality to the research community, we hope to contribute to a better understanding of intrinsic dimensionality and to spur new research in this direction. △ Less

Submitted 23 June, 2020; originally announced June 2020.

ELKI: A large open-source library for data analysis - ELKI Release 0.7.5 "Heidelberg"

Authors: Erich Schubert, Arthur Zimek

Abstract: This paper documents the release of the ELKI data mining framework, version 0.7.5. ELKI is an open source (AGPLv3) data mining software written in Java. The focus of ELKI is research in algorithms, with an emphasis on unsupervised methods in cluster analysis and outlier detection. In order to achieve high performance and scalability, ELKI offers data index structures such as the R*-tree that can… ▽ More This paper documents the release of the ELKI data mining framework, version 0.7.5. ELKI is an open source (AGPLv3) data mining software written in Java. The focus of ELKI is research in algorithms, with an emphasis on unsupervised methods in cluster analysis and outlier detection. In order to achieve high performance and scalability, ELKI offers data index structures such as the R*-tree that can provide major performance gains. ELKI is designed to be easy to extend for researchers and students in this domain, and welcomes contributions of additional methods. ELKI aims at providing a large collection of highly parameterizable algorithms, in order to allow easy and fair evaluation and benchmarking of algorithms. We will first outline the motivation for this release, the plans for the future, and then give a brief overview over the new functionality in this version. We also include an appendix presenting an overview on the overall implemented functionality. △ Less

Submitted 10 February, 2019; originally announced February 2019.

Faster k-Medoids Clustering: Improving the PAM, CLARA, and CLARANS Algorithms

Authors: Erich Schubert, Peter J. Rousseeuw

Abstract: Clustering non-Euclidean data is difficult, and one of the most used algorithms besides hierarchical clustering is the popular algorithm Partitioning Around Medoids (PAM), also simply referred to as k-medoids. In Euclidean geometry the mean-as used in k-means-is a good estimator for the cluster center, but this does not hold for arbitrary dissimilarities. PAM uses the medoid instead, the object wi… ▽ More Clustering non-Euclidean data is difficult, and one of the most used algorithms besides hierarchical clustering is the popular algorithm Partitioning Around Medoids (PAM), also simply referred to as k-medoids. In Euclidean geometry the mean-as used in k-means-is a good estimator for the cluster center, but this does not hold for arbitrary dissimilarities. PAM uses the medoid instead, the object with the smallest dissimilarity to all others in the cluster. This notion of centrality can be used with any (dis-)similarity, and thus is of high relevance to many domains such as biology that require the use of Jaccard, Gower, or more complex distances. A key issue with PAM is its high run time cost. We propose modifications to the PAM algorithm to achieve an O(k)-fold speedup in the second SWAP phase of the algorithm, but will still find the same results as the original PAM algorithm. If we slightly relax the choice of swaps performed (at comparable quality), we can further accelerate the algorithm by performing up to k swaps in each iteration. With the substantially faster SWAP, we can now also explore alternative strategies for choosing the initial medoids. We also show how the CLARA and CLARANS algorithms benefit from these modifications. It can easily be combined with earlier approaches to use PAM and CLARA on big data (some of which use PAM as a subroutine, hence can immediately benefit from these improvements), where the performance with high k becomes increasingly important. In experiments on real data with k=100, we observed a 200-fold speedup compared to the original PAM SWAP algorithm, making PAM applicable to larger data sets as long as we can afford to compute a distance matrix, and in particular to higher k (at k=2, the new SWAP was only 1.5 times faster, as the speedup is expected to increase with k). △ Less

Submitted 29 October, 2019; v1 submitted 12 October, 2018; originally announced October 2018.

Journal ref: Similarity Search and Applications, SISAP 2019

Tunable plasmonic resonances in highly porous nano-bamboo Si-Au superlattice-type thin films

Authors: Ufuk Kilic, Alyssa Mock, Rene Feder, Derek Sekora, Matthew Hilfiker, Rafal Korlacki, Eva Schubert, Christos Argyropoulos, Mathias Schubert

Abstract: We report on fabrication of spatially-coherent columnar plasmonic nanostructure superlattice-type thin films with high porosity and strong optical anisotropy using glancing angle deposition. Subsequent and repeated depositions of silicon and gold lead to nanometer-dimension subcolumns with controlled lengths. The superlattice-type columns resemble bamboo structures where smaller column sections of… ▽ More We report on fabrication of spatially-coherent columnar plasmonic nanostructure superlattice-type thin films with high porosity and strong optical anisotropy using glancing angle deposition. Subsequent and repeated depositions of silicon and gold lead to nanometer-dimension subcolumns with controlled lengths. The superlattice-type columns resemble bamboo structures where smaller column sections of gold form junctions sandwiched between larger silicon column sections ("nano-bamboo"). We perform generalized spectroscopic ellipsometry measurements and finite element method computations to elucidate the strongly anisotropic optical properties of the highly-porous nano-bamboo structures. The occurrence of a strongly localized plasmonic mode with displacement pattern reminiscent of a dark quadrupole mode is observed in the vicinity of the gold subcolumns. We demonstrate tuning of this quadrupole-like mode frequency within the near-infrared spectral range by varying the geometry of the nano-bamboo structure. In addition, coupled-plasmon-like and inter-band transition-like modes occur in the visible and ultra-violet spectral regions, respectively. We elucidate an example for the potential use of the nano-bamboo structures as a highly porous plasmonic sensor with optical read out sensitivity to few parts-per-million solvent levels in water. △ Less

Submitted 16 July, 2018; originally announced July 2018.

An Improved Method of Total Variation Superiorization Applied to Reconstruction in Proton Computed Tomography

Authors: Blake Schultze, Yair Censor, Paniz Karbasi, Keith E. Schubert, Reinhard W. Schulte

Abstract: Previous work showed that total variation superiorization (TVS) improves reconstructed image quality in proton computed tomography (pCT). The structure of the TVS algorithm has evolved since then and this work investigated if this new algorithmic structure provides additional benefits to pCT image quality. Structural and parametric changes introduced to the original TVS algorithm included: (1) inc… ▽ More Previous work showed that total variation superiorization (TVS) improves reconstructed image quality in proton computed tomography (pCT). The structure of the TVS algorithm has evolved since then and this work investigated if this new algorithmic structure provides additional benefits to pCT image quality. Structural and parametric changes introduced to the original TVS algorithm included: (1) inclusion or exclusion of TV reduction requirement, (2) a variable number, $N$, of TV perturbation steps per feasibility-seeking iteration, and (3) introduction of a perturbation kernel $0<α<1$. The structural change of excluding the TV reduction requirement check tended to have a beneficial effect for $3\le N\le 6$ and allows full parallelization of the TVS algorithm. Repeated perturbations per feasibility-seeking iterations reduced total variation (TV) and material dependent standard deviations for $3\le N\le 6$. The perturbation kernel $α$, equivalent to $α=0.5$ in the original TVS algorithm, reduced TV and standard deviations as $α$ was increased beyond $α=0.5$, but negatively impacted reconstructed relative stop** power (RSP) values for $α>0.75$. The reductions in TV and standard deviations allowed feasibility-seeking with a larger relaxation parameter $λ$ than previously used, without the corresponding increases in standard deviations experienced with the original TVS algorithm. This work demonstrates that the modifications related to the evolution of the original TVS algorithm provide benefits in terms of both pCT image quality and computational efficiency for appropriately chosen parameter values. △ Less

Submitted 17 January, 2019; v1 submitted 3 March, 2018; originally announced March 2018.

A Highly Accelerated Parallel Multi-GPU based Reconstruction Algorithm for Generating Accurate Relative Stop** Powers

Authors: Paniz Karbasi, Ritchie Cai, Blake Schultze, Hanh Nguyen, Jones Reed, Patrick Hall, Valentina Giacometti, Vladimir Bashkirov, Robert Johnson, Nick Karonis, Jeffrey Olafsen, Caesar Ordonez, Keith E. Schubert, Reinhard W. Schulte

Abstract: Low-dose Proton Computed Tomography (pCT) is an evolving imaging modality that is used in proton therapy planning which addresses the range uncertainty problem. The goal of pCT is generating a 3D map of Relative Stop** Power (RSP) measurements with high accuracy within clinically required time frames. Generating accurate RSP values within the shortest amount of time is considered a key goal when… ▽ More Low-dose Proton Computed Tomography (pCT) is an evolving imaging modality that is used in proton therapy planning which addresses the range uncertainty problem. The goal of pCT is generating a 3D map of Relative Stop** Power (RSP) measurements with high accuracy within clinically required time frames. Generating accurate RSP values within the shortest amount of time is considered a key goal when develo** a pCT software. The existing pCT softwares have successfully met this time frame and even succeeded this time goal, but requiring clusters with hundreds of processors. This paper describes a novel reconstruction technique using two Graphics Processing Unit (GPU) cores, such as is available on a single Nvidia P100. The proposed reconstruction technique is tested on both simulated and experimental datasets and on two different systems namely Nvidia K40 and P100 GPUs from IBM and Cray. The experimental results demonstrate that our proposed reconstruction method meets both the timing and accuracy with the benefit of having reasonable cost, and efficient use of power. △ Less

Submitted 3 February, 2018; originally announced February 2018.

Comments: IEEE NSS/MIC 2017

HV discharge acceleration by sequences of UV laser filaments with visible and near-infrared pulses

Authors: Elise Schubert, Ali Rastegari, Chengyong Feng, Denis Mongin, Brian Kamer, Jérôme Kasparian, Jean-Pierre Wolf, Ladan Arissian, Jean-Claude Diels

Abstract: We investigate the triggering and guiding of DC high-voltage discharges over a distance of 37 cm by filaments produced by ultraviolet (266 nm) laser pulses of 200 ps duration. The latter reduce the breakdown electric field by half and allow up to 80% discharge probability in an electric field of 920 kV/m. This high efficiency is not further increased by adding nanosecond pulses in the Joule range… ▽ More We investigate the triggering and guiding of DC high-voltage discharges over a distance of 37 cm by filaments produced by ultraviolet (266 nm) laser pulses of 200 ps duration. The latter reduce the breakdown electric field by half and allow up to 80% discharge probability in an electric field of 920 kV/m. This high efficiency is not further increased by adding nanosecond pulses in the Joule range at 532 nm and 1064 nm. However, the latter statistically increases the guiding length, thereby accelerating the discharge by a factor of 2. This effect is due both to photodetachment and to the heating of the plasma channel, that increases the efficiency of avalanche ionization and reduces electron attachment and recombination. △ Less

Submitted 22 December, 2017; originally announced December 2017.

Comments: 12 pages, 6 figures

Journal ref: New Journal of Physics 19, 123040 (2017)

Linearity of charge measurement in laser filaments

Authors: Denis Mongin, Elise Schubert, Lorena de la Cruz, Nicolas Berti, Jérôme Kasparian, Jean-Pierre Wolf

Abstract: We evaluate the linearity of three electric measurement techniques of the initial electron density in laser filaments by comparing their results for a pair of filaments and for the sum of each individual filament. The conductivity measured between two plane electrodes in a longitudinal configuration is linear within 2% provided the electric field is kept below 100 kV/m. Furthermore, simulations sh… ▽ More We evaluate the linearity of three electric measurement techniques of the initial electron density in laser filaments by comparing their results for a pair of filaments and for the sum of each individual filament. The conductivity measured between two plane electrodes in a longitudinal configuration is linear within 2% provided the electric field is kept below 100 kV/m. Furthermore, simulations show that the signal behaves like the amount of generated free electrons. The slow ionic current measured with plane electrodes in a parallel configuration is representative of the ionic charge available in the filament, after several $μ$s, when the free electrons have recombined. It is linear within 2% with the amount of ions and is insensitive to misalignment. Finally, the fast polarization signal in the same configuration deviates from linearity by up to 80% and can only be considered as a semi-qualitative indication of the presence of charges, e.g., to characterize the filament length. △ Less

Submitted 3 July, 2017; originally announced August 2017.

Comments: 17 pages, 7 figures

Journal ref: Optics Express Vol. 25, Issue 14, pp. 16517-16526 (2017)

Semantic Word Clouds with Background Corpus Normalization and t-distributed Stochastic Neighbor Embedding

Authors: Erich Schubert, Andreas Spitz, Michael Weiler, Johanna Geiß, Michael Gertz

Abstract: Many word clouds provide no semantics to the word placement, but use a random layout optimized solely for aesthetic purposes. We propose a novel approach to model word significance and word affinity within a document, and in comparison to a large background corpus. We demonstrate its usefulness for generating more meaningful word clouds as a visual summary of a given document. We then select keywo… ▽ More Many word clouds provide no semantics to the word placement, but use a random layout optimized solely for aesthetic purposes. We propose a novel approach to model word significance and word affinity within a document, and in comparison to a large background corpus. We demonstrate its usefulness for generating more meaningful word clouds as a visual summary of a given document. We then select keywords based on their significance and construct the word cloud based on the derived affinity. Based on a modified t-distributed stochastic neighbor embedding (t-SNE), we generate a semantic word placement. For words that cooccur significantly, we include edges, and cluster the words according to their cooccurrence. For this we designed a scalable and memory-efficient sketch-based approach usable on commodity hardware to aggregate the required corpus statistics needed for normalization, and for identifying keywords as well as significant cooccurences. We empirically validate our approch using a large Wikipedia corpus. △ Less

Submitted 11 August, 2017; originally announced August 2017.

Results from a Prototype Proton-CT Head Scanner

Authors: R. P. Johnson, V. A. Bashkirov, G. Coutrakon, V. Giacometti, P. Karbasi, N. T. Karonis, C. E. Ordoñez, M. Pankuch, H. F. -W. Sadrozinski, K. E. Schubert, R. W. Schulte

Abstract: We are exploring low-dose proton radiography and computed tomography (pCT) as techniques to improve the accuracy of proton treatment planning and to provide artifact-free images for verification and adaptive therapy at the time of treatment. Here we report on comprehensive beam test results with our prototype pCT head scanner. The detector system and data acquisition attain a sustained rate of mor… ▽ More We are exploring low-dose proton radiography and computed tomography (pCT) as techniques to improve the accuracy of proton treatment planning and to provide artifact-free images for verification and adaptive therapy at the time of treatment. Here we report on comprehensive beam test results with our prototype pCT head scanner. The detector system and data acquisition attain a sustained rate of more than a million protons individually measured per second, allowing a full CT scan to be completed in six minutes or less of beam time. In order to assess the performance of the scanner for proton radiography as well as computed tomography, we have performed numerous scans of phantoms at the Northwestern Medicine Chicago Proton Center including a custom phantom designed to assess the spatial resolution, a phantom to assess the measurement of relative stop** power, and a dosimetry phantom. Some images, performance, and dosimetry results from those phantom scans are presented together with a description of the instrument, the data acquisition system, and the calibration methods. △ Less

Submitted 5 July, 2017; originally announced July 2017.

Comments: Conference on the Application of Accelerators in Research and Industry, CAARI 2016, 30 October to 4 November 2016, Ft. Worth, TX, USA

Gas-solid phase transition in laser multiple filamentation

Authors: Denis Mongin, Elise Schubert, Nicolas Berti, Jérôme Kasparian, Jean-Pierre Wolf

Abstract: While propagating in transparent media, near-infrared multi-terawatt (TW) laser beams break up in a multitude of filaments of typically 100-200 um diameter with peak intensities as high as 10 to 100~TW/cm$^{2}$. We observe a phase transition at incident beam intensities of 0.4~TW/cm$^2$, where the interaction between filaments induce solid-like 2-dimensional crystals with a 2.7 mm lattice constant… ▽ More While propagating in transparent media, near-infrared multi-terawatt (TW) laser beams break up in a multitude of filaments of typically 100-200 um diameter with peak intensities as high as 10 to 100~TW/cm$^{2}$. We observe a phase transition at incident beam intensities of 0.4~TW/cm$^2$, where the interaction between filaments induce solid-like 2-dimensional crystals with a 2.7 mm lattice constant, independent of the initial beam diameter. Below 0.4~TW/cm$^2$, we evidence a mixed phase state in which some filaments are closely packed in localized clusters, nucleated on inhomogeneities (seeds) in the transverse intensity profile of the beam, and other are sparse with almost no interaction with their neighbors, similar to a gas. This analogy with a thermodynamic gas-solid phase transition is confirmed by calculating the interaction Hamiltonian between neighboring filaments, which takes into account the effect of diffraction, Kerr self-focusing and plasma generation. The shape of the effective potential is close to a Morse potential with an equilibrium bond length close to the observed value. △ Less

Submitted 13 April, 2017; originally announced April 2017.

Journal ref: Phys. Rev. Lett. 118, 133902 (2017)

High repetition rate ultrashort laser cuts a path through fog

Authors: Lorena de la Cruz, Elise Schubert, Denis Mongin, Sandro Klingebiel, Marcel Schultze, Thomas Metzger, Knut Michel, Jérôme Kasparian, Jean-Pierre Wolf

Abstract: We experimentally demonstrate that the transmission of a 1030~nm, 1.3~ps laser beam of 100 mJ energy through fog increases when its repetition rate increases to the kHz range. Due to the efficient energy deposition by the laser filaments in the air, a shockwave ejects the fog droplets from a substantial volume of the beam, at a moderate energy cost. This process opens prospects for applications re… ▽ More We experimentally demonstrate that the transmission of a 1030~nm, 1.3~ps laser beam of 100 mJ energy through fog increases when its repetition rate increases to the kHz range. Due to the efficient energy deposition by the laser filaments in the air, a shockwave ejects the fog droplets from a substantial volume of the beam, at a moderate energy cost. This process opens prospects for applications requiring the transmission of laser beams through fogs and clouds. △ Less

Submitted 25 December, 2016; originally announced December 2016.

Comments: 11 pages, 5 figures

Journal ref: Applied Physics Letters 109, 251105 (2016)

Conductivity and Discharge Guiding Properties of Mid-IR Laser Filaments

Authors: Denis Mongin, Valentina Shumakova, Skirmantas Ališauskas, Audrius Pugzlys, Elise Schubert, jerome kasparian, Jean Pierre Wolf, Andrius Baltuska

Abstract: The electric conductivity, HV discharge triggering and guiding capabilities of filaments at 3.9 micrometer in air are investigated in the perspective of lightning control applications, and compared to near-IR filaments in identical conditions The electric conductivity, HV discharge triggering and guiding capabilities of filaments at 3.9 micrometer in air are investigated in the perspective of lightning control applications, and compared to near-IR filaments in identical conditions △ Less

Submitted 4 November, 2016; originally announced November 2016.

Journal ref: Appl. Phys. B 122: 267 (2016)

Anisotropic Contrast Optical Microscope

Authors: D. Peev, T. Hofmann, N. Kananizadeh, S. Wimer, K. B. Rodenhausen, C. M. Herzinger, T. Kasputis, E. Pfaunmiller, A. Nguyen, R. Korlacki, A. Pannier, Y. Li, E. Schubert, D. Hage, M. Schubert

Abstract: An optical microscope is described that reveals contrast in the Mueller matrix images of a thin, transparent or semi-transparent specimen located within an anisotropic object plane (anisotropic filter). The specimen changes the anisotropy of the filter and thereby produces contrast within the Mueller matrix images. Here we use an anisotropic filter composed of a semi-transparent, nanostructured th… ▽ More An optical microscope is described that reveals contrast in the Mueller matrix images of a thin, transparent or semi-transparent specimen located within an anisotropic object plane (anisotropic filter). The specimen changes the anisotropy of the filter and thereby produces contrast within the Mueller matrix images. Here we use an anisotropic filter composed of a semi-transparent, nanostructured thin film with sub-wavelength thickness placed within the object plane. The sample is illuminated as in common optical microscopy but the light is modulated in its polarization using combinations of linear polarizers and phase plate (compensator) to control and analyze the state of polarization. Direct generalized ellipsometry data analysis approaches permit extraction of fundamental Mueller matrix object plane images dispensing with the need of Fourier expansion methods. Generalized ellipsometry model approaches are used for quantitative image analyses. We demonstrate the anisotropic contrast optical microscope by measuring lithographically defined micro-patterned anisotropic filters, and we quantify the adsorption of an organic self-assembled monolayer film onto the anisotropic filter. In our current instrument we estimate the limit of detection for organic volumetric mass within the object plane of $\approx$ 49 fg within $\approx$ 7$\times$7~$μ$m$^2$ object surface area. Compared to a quartz crystal microbalance with dissipation instrumentation, where contemporary limits require a total load of $\approx$ 500~pg for detection, the instrumentation demonstrated here improves sensitivity to a total mass required for detection by 4 orders of magnitude. We present further applications to detection of nanoparticles, to novel approaches for imaging chromatography, and to new contrast modalities for observations on living cells. △ Less

Submitted 14 October, 2016; originally announced October 2016.

Comments: 18 pages, 13 figures, accepted as regular article in Review of Scientific Instruments (October 2016)

Journal ref: Review of Scientific Instruments 87, 113701 (2016)

Dual-scale turbulence in filamenting laser beams at high average power

Authors: Elise Schubert, Lorena de la Cruz, Denis Mongin, Jérôme Kasparian, Jean-Pierre Wolf, Sandro Klingebiel, Marcel Schultze, Thomas Metzger, Knut Michel

Abstract: We investigate the self-induced turbulence of high repetition rate laser filaments over a wide range of average powers (1 mW to 100 W) and its sensitivity to external atmospheric turbulence. Although both externally-imposed and self-generated turbulences can have comparable magnitudes, they act on different temporal and spatial scales. While the former drives the shot-to-shot motion at the millise… ▽ More We investigate the self-induced turbulence of high repetition rate laser filaments over a wide range of average powers (1 mW to 100 W) and its sensitivity to external atmospheric turbulence. Although both externally-imposed and self-generated turbulences can have comparable magnitudes, they act on different temporal and spatial scales. While the former drives the shot-to-shot motion at the millisecond time scale, the latter acts on the 0.5 s scale. As a consequence, their effects are decoupled, preventing beam stabilization by the thermally-induced low-density channel produced by the laser filaments. △ Less

Submitted 11 October, 2016; originally announced October 2016.

Comments: 11 pages, 5 figures

Journal ref: Physical Review A 94, 043808 (2016)

Optimal laser pulse energy partitioning for air ionization

Authors: Elise Schubert, Jean-Gabriel Brisset, Mary Matthews, Antoine Courjaud, Jérôme Kasparian, Jean-Pierre Wolf

Abstract: We investigate the pulse partitioning of a 6.3 mJ, 450 fs pulse at 1030 nm to produce plasma channels. At such moderate energies, splitting the energy into several sub-pulses reduces the ionization efficiency and thus does not extend the plasma lifetime. We numerically show that when sufficient energy to produce multifilamentation is available, splitting the pulse temporally in a pulse train incre… ▽ More We investigate the pulse partitioning of a 6.3 mJ, 450 fs pulse at 1030 nm to produce plasma channels. At such moderate energies, splitting the energy into several sub-pulses reduces the ionization efficiency and thus does not extend the plasma lifetime. We numerically show that when sufficient energy to produce multifilamentation is available, splitting the pulse temporally in a pulse train increases the gas temperature compared to a filament bundle of the same energy. This could improve the mean free path of the free electrons, therefore enhancing the efficiency of discharge triggering. △ Less

Submitted 20 December, 2022; v1 submitted 24 August, 2016; originally announced August 2016.

Comments: Correcting typo in Eq. (12) and unit of $β_{ep}$ in Table I

Journal ref: Phys. Rev. A 94, 033824 (2016)

Observation of dispersive shock waves, solitons, and their interactions in viscous fluid conduits

Authors: Michelle D. Maiden, Nicholas K. Lowman, Dalton V. Anderson, Marika E. Schubert, Mark A. Hoefer

Abstract: Dispersive shock waves and solitons are fundamental nonlinear excitations in dispersive media, but dispersive shock wave studies to date have been severely constrained. Here we report on a novel dispersive hydrodynamics testbed: the effectively frictionless dynamics of interfacial waves between two high contrast, miscible, low Reynolds' number Stokes fluids. This scenario is realized by injecting… ▽ More Dispersive shock waves and solitons are fundamental nonlinear excitations in dispersive media, but dispersive shock wave studies to date have been severely constrained. Here we report on a novel dispersive hydrodynamics testbed: the effectively frictionless dynamics of interfacial waves between two high contrast, miscible, low Reynolds' number Stokes fluids. This scenario is realized by injecting from below a lighter, viscous fluid into a column filled with high viscosity fluid. The injected fluid forms a deformable pipe whose diameter is proportional to the injection rate, enabling precise control over the generation of symmetric interfacial waves. Buoyancy drives nonlinear interfacial self-steepening while normal stresses give rise to dispersion of interfacial waves. Extremely slow mass diffusion and mass conservation imply that the interfacial waves are effectively dissipationless. This enables high fidelity observations of large amplitude dispersive shock waves in this spatially extended system, found to agree quantitatively with a nonlinear wave averaging theory. Furthermore, several highly coherent phenomena are investigated including dispersive shock wave backflow, the refraction or absorption of solitons by dispersive shock waves, and the multi-phase merging of two dispersive shock waves. The complex, coherent, nonlinear mixing of dispersive shock waves and solitons observed here are universal features of dissipationless, dispersive hydrodynamic flows. △ Less

Submitted 12 April, 2016; v1 submitted 31 December, 2015; originally announced December 2015.

Comments: 10 pages, 9 figures

Journal ref: Phys. Rev. Lett. 116, 174501 (2016)

Anisotropy, band-to-band transitions, phonon modes, and oxidation properties of cobalt-oxide core-shell slanted columnar thin films

Authors: Alyssa Mock, Rafal Korlacki, Chad Briley, Derek Sekora, Tino Hofmann, Peter Wilson, Alexander Sinitskii, Eva Schubert, Mathias Schubert

Abstract: Highly-ordered and spatially-coherent cobalt slanted columnar thin films were deposited by glancing angle deposition onto silicon substrates, and subsequently oxidized by annealing at 475 $^{\circ}$C. Scanning electron microscopy, Raman scattering, generalized ellipsometry, and density functional theory investigations reveal shape-invariant transformation of the slanted nanocolumns from metallic t… ▽ More Highly-ordered and spatially-coherent cobalt slanted columnar thin films were deposited by glancing angle deposition onto silicon substrates, and subsequently oxidized by annealing at 475 $^{\circ}$C. Scanning electron microscopy, Raman scattering, generalized ellipsometry, and density functional theory investigations reveal shape-invariant transformation of the slanted nanocolumns from metallic to transparent metal-oxide core-shell structures with properties characteristic of spinel cobalt oxide. We find passivation of Co-SCTFs yielding CoAl$_2$O$_3$ core-shell structures produced by conformal deposition of a few nanometers of alumina using atomic layer deposition fully prevents cobalt oxidation in ambient and from annealing up to 475 $^{\circ}$C. △ Less

Submitted 6 December, 2015; originally announced December 2015.

Comments: 5 pages, 5 figures, 2 tables

Non-linear photochemical pathways in laser induced atmospheric aerosol formation

Authors: Denis Mongin, Jay G. Slowik, Elise Schubert, Jean-Gabriel Brisset, Nicolas Berti, Michel Moret, André S. H. Prévôt, Urs Baltensperger, Jérôme Kasparian, Jean-Pierre Wolf

Abstract: We measured the chemical composition and the size distribution of aerosols generated by femtosecond-Terawatt laser pulses in the atmosphere using an aerosol mass spectrometer (AMS). We show that nitric acid condenses in the form of ammonium nitrate, and that oxidized volatile organics also contribute to particle growth. These two components account for two thirds and one third, respectively, of th… ▽ More We measured the chemical composition and the size distribution of aerosols generated by femtosecond-Terawatt laser pulses in the atmosphere using an aerosol mass spectrometer (AMS). We show that nitric acid condenses in the form of ammonium nitrate, and that oxidized volatile organics also contribute to particle growth. These two components account for two thirds and one third, respectively, of the dry laser-condensed mass. They appear in two different modes centred at 380 nm and 150 nm. The number concentration of particles between 25 and 300 nm increases by a factor of 15. Pre-existing water droplets strongly increase the oxidative properties of the laser-activated atmosphere, substantially enhancing the condensation of organics under laser illumination. △ Less

Submitted 15 September, 2015; originally announced September 2015.

Comments: 19 pages, 5 figures

Remote electrical arc suppression by laser filamentation

Authors: Elise Schubert, Denis Mongin, Jérôme Kasparian, Jean-Pierre Wolf

Abstract: We investigate the interaction of narrow plasma channels formed in the filamentation of ultrashort laser pulses, with a DC high voltage. The laser filaments prevent electrical arcs by triggering corona that neutralize the high-voltage electrodes. This phenomenon, due to the electric field modulation and free electron release around the filament, opens new prospects to lightning and over-voltage mi… ▽ More We investigate the interaction of narrow plasma channels formed in the filamentation of ultrashort laser pulses, with a DC high voltage. The laser filaments prevent electrical arcs by triggering corona that neutralize the high-voltage electrodes. This phenomenon, due to the electric field modulation and free electron release around the filament, opens new prospects to lightning and over-voltage mitigation. △ Less

Submitted 5 May, 2017; v1 submitted 24 August, 2015; originally announced August 2015.

Comments: 15 pages, 6 figures

Journal ref: Optics Express Vol. 23, Issue 22, pp. 28640-28648 (2015); Erratum in Optics Express vol. 25, issue 10, pp. 11210 (2017)

Confocal shift interferometry of coherent emission from trapped dipolar excitons

Authors: Jens Repp, Georg J. Schinner, Enrico Schubert, Ashish K. Rai, Dirk Reuter, Andreas D. Wieck, Ursula Wurstbauer, Joerg P. Kotthaus, Alexander W. Holleitner

Abstract: We introduce a confocal shift-interferometer based on optical fibers. The presented spectroscopy allows measuring coherence maps of luminescent samples with a high spatial resolution even at cryogenic temperatures. We apply the spectroscopy onto electrostatically trapped, dipolar excitons in a semiconductor double quantum well. We find that the measured spatial coherence length of the excitonic em… ▽ More We introduce a confocal shift-interferometer based on optical fibers. The presented spectroscopy allows measuring coherence maps of luminescent samples with a high spatial resolution even at cryogenic temperatures. We apply the spectroscopy onto electrostatically trapped, dipolar excitons in a semiconductor double quantum well. We find that the measured spatial coherence length of the excitonic emission coincides with the point spread function of the confocal setup. The results are consistent with a temporal coherence of the excitonic emission down to temperatures of 250 mK. △ Less

Submitted 19 November, 2014; originally announced November 2014.

Comments: 5 pages, 3 figures

On the relation between the 0.7-anomaly and the Kondo effect: Geometric Crossover between a Quantum Point Contact and a Kondo Quantum Dot

Authors: Jan Heyder, Florian Bauer, Enrico Schubert, David Borowsky, Dieter Schuh, Werner Wegscheider, Jan von Delft, Stefan Ludwig

Abstract: Quantum point contacts (QPCs) and quantum dots (QDs), two elementary building blocks of semiconducting nanodevices, both exhibit famously anomalous conductance features: the 0.7-anomaly in the former case, the Kondo effect in the latter. For both the 0.7-anomaly and the Kondo effect, the conductance shows a remarkably similar low-energy dependence on temperature $T$, source-drain voltage… ▽ More Quantum point contacts (QPCs) and quantum dots (QDs), two elementary building blocks of semiconducting nanodevices, both exhibit famously anomalous conductance features: the 0.7-anomaly in the former case, the Kondo effect in the latter. For both the 0.7-anomaly and the Kondo effect, the conductance shows a remarkably similar low-energy dependence on temperature $T$, source-drain voltage $V_{\rm sd}$ and magnetic field $B$. In a recent publication [F. Bauer et al., Nature, 501, 73 (2013)], we argued that the reason for these similarities is that both a QPC and a KQD feature spin fluctuations that are induced by the sample geometry, confined in a small spatial regime, and enhanced by interactions. Here we further explore this notion experimentally and theoretically by studying the geometric crossover between a QD and a QPC, focussing on the $B$-field dependence of the conductance. We introduce a one-dimensional model that reproduces the essential features of the experiments, including a smooth transition between a Kondo QD and a QPC with 0.7-anomaly. We find that in both cases the anomalously strong negative magnetoconductance goes hand in hand with strongly enhanced local spin fluctuations. Our experimental observations include, in addition to the Kondo effect in a QD and the 0.7-anomaly in a QPC, Fano interference effects in a regime of coexistence between QD and QPC physics, and Fabry-Perot-type resonances on the conductance plateaus of a clean QPC. We argue that Fabry-Perot-type resonances occur generically if the electrostatic potential of the QPC generates a flatter-than-parabolic barrier top. △ Less

Submitted 11 September, 2014; originally announced September 2014.

Journal ref: Phys. Rev. B 92, 195401 (2015)

Sub-Kelvin optical thermometry of an electron reservoir coupled to a self-assembled InGaAs quantum dot

Authors: F. Seilmeier, M. Hauck, E. Schubert, G. J. Schinner, S. E. Beavan, A. Högele

Abstract: We show how resonant laser spectroscopy of the trion optical transitions in a self-assembled quantum dot can be used to determine the temperature of a nearby electron reservoir. At finite magnetic field the spin-state occupation of the Zeeman-split quantum dot electron ground states is governed by thermalization with the electron reservoir via co-tunneling. With resonant spectroscopy of the corres… ▽ More We show how resonant laser spectroscopy of the trion optical transitions in a self-assembled quantum dot can be used to determine the temperature of a nearby electron reservoir. At finite magnetic field the spin-state occupation of the Zeeman-split quantum dot electron ground states is governed by thermalization with the electron reservoir via co-tunneling. With resonant spectroscopy of the corresponding excited trion states we map out the spin occupation as a function of magnetic field to establish optical thermometry for the electron reservoir. We demonstrate the implementation of the technique in the sub-Kelvin temperature range where it is most sensitive, and where the electron temperature is not necessarily given by the cryostat base temperature. △ Less

Submitted 9 May, 2014; originally announced May 2014.

Comments: 5 pages, 4 figures

Journal ref: Phys. Rev. Applied 2, 024002 (2014)

Towards combined transport and optical studies of the 0.7-anomaly in a quantum point contact

Authors: E. Schubert, J. Heyder, F. Bauer, W. Stumpf, W. Wegscheider, J. v. Delft, S. Ludwig, A. Högele

Abstract: A Quantum Point Contact (QPC) causes a one-dimensional constriction on the spatial potential landscape of a two-dimensional electron system. By tuning the voltage applied on a QPC at low temperatures the resulting regular step-like electron conductance quantization can show an additional kink near pinch-off around 0.7(2$e^2$/h), called 0.7-anomaly. In a recent publication, we presented a combinati… ▽ More A Quantum Point Contact (QPC) causes a one-dimensional constriction on the spatial potential landscape of a two-dimensional electron system. By tuning the voltage applied on a QPC at low temperatures the resulting regular step-like electron conductance quantization can show an additional kink near pinch-off around 0.7(2$e^2$/h), called 0.7-anomaly. In a recent publication, we presented a combination of theoretical calculations and transport measurements that lead to a detailed understanding of the microscopic origin of the 0.7-anomaly. Functional Renormalization Group-based calculations were performed exhibiting the 0.7-anomaly even when no symmetry-breaking external magnetic fields are involved. According to the calculations the electron spin susceptibility is enhanced within a QPC that is tuned in the region of the 0.7-anomaly. Moderate externally applied magnetic fields impose a corresponding enhancement in the spin magnetization. In principle, it should be possible to map out this spin distribution optically by means of the Faraday rotation technique. Here we report the initial steps of an experimental project aimed at realizing such measurements. Simulations were performed on a particularly pre-designed semiconductor heterostructure. Based on the simulation results a sample was built and its basic transport and optical properties were investigated. Finally, we introduce a sample gate design, suitable for combined transport and optical studies. △ Less

Submitted 31 March, 2014; originally announced March 2014.

Performance of Hull-Detection Algorithms For Proton Computed Tomography Reconstruction

Authors: Blake Schultze, Micah Witt, Yair Censor, Reinhard Schulte, Keith Evan Schubert

Abstract: Proton computed tomography (pCT) is a novel imaging modality developed for patients receiving proton radiation therapy. The purpose of this work was to investigate hull-detection algorithms used for preconditioning of the large and sparse linear system of equations that needs to be solved for pCT image reconstruction. The hull-detection algorithms investigated here included silhouette/space carvin… ▽ More Proton computed tomography (pCT) is a novel imaging modality developed for patients receiving proton radiation therapy. The purpose of this work was to investigate hull-detection algorithms used for preconditioning of the large and sparse linear system of equations that needs to be solved for pCT image reconstruction. The hull-detection algorithms investigated here included silhouette/space carving (SC), modified silhouette/space carving (MSC), and space modeling (SM). Each was compared to the cone-beam version of filtered backprojection (FBP) used for hull-detection. Data for testing these algorithms included simulated data sets of a digital head phantom and an experimental data set of a pediatric head phantom obtained with a pCT scanner prototype at Loma Linda University Medical Center. SC was the fastest algorithm, exceeding the speed of FBP by more than 100 times. FBP was most sensitive to the presence of noise. Ongoing work will focus on optimizing threshold parameters in order to define a fast and efficient method for hull-detection in pCT image reconstruction. △ Less

Submitted 7 February, 2014; originally announced February 2014.

Comments: Contemporary Mathematics, accepted for publication

Structural and optical properties of cobalt slanted columnar thin films conformally coated with graphene by chemical vapor deposition

Authors: P. M. Wilson, D. Schmidt, E. Schubert, M. Schubert, A. Sinitskii, T. Hofmann

Abstract: A slanted cobalt sculptured columnar thin film was fabricated using glancing angle deposition, and coated subsequently with graphene using a low temperature chemical vapor deposition process. The graphene deposition process preserves shape and geometry of the sculptured thin film, which was confirmed by scanning electron microscopy. According to the Raman spectroscopy results, the graphene coating… ▽ More A slanted cobalt sculptured columnar thin film was fabricated using glancing angle deposition, and coated subsequently with graphene using a low temperature chemical vapor deposition process. The graphene deposition process preserves shape and geometry of the sculptured thin film, which was confirmed by scanning electron microscopy. According to the Raman spectroscopy results, the graphene coating is two to three monolayers thick and has a high defect concentration. The graphene coverage within the sculptured thin film is determined from generalized spectroscopic ellipsometry using a generalized anisotropic Bruggeman effective medium approximation. The graphene coverage as well as structural parameters of the thin film agree excellently with electron microscopy and Raman observations, and suggest that the graphene coating is conformal. △ Less

Submitted 18 December, 2013; originally announced December 2013.

Single exciton emission from gate-defined quantum dots

Authors: G. J. Schinner, J. Repp, E. Schubert, A. K. Rai, D. Reuter, A. D. Wieck, A. O. Govorov, A. W. Holleitner, J. P. Kotthaus

Abstract: With gate-defined electrostatic traps fabricated on a double quantum well we are able to realize an optically active and voltage-tunable quantum dot confining individual, long-living, spatially indirect excitons. We study the transition from multi excitons down to a single indirect exciton. In the few exciton regime, we observe discrete emission lines reflecting the interplay of dipolar interexcit… ▽ More With gate-defined electrostatic traps fabricated on a double quantum well we are able to realize an optically active and voltage-tunable quantum dot confining individual, long-living, spatially indirect excitons. We study the transition from multi excitons down to a single indirect exciton. In the few exciton regime, we observe discrete emission lines reflecting the interplay of dipolar interexcitonic repulsion and spatial quantization. The quantum dot states are tunable by gate voltage and employing a magnetic field results in a diamagnetic shift. The scheme introduces a new gate-defined platform for creating and controlling optically active quantum dots and opens the route to lithographically defined coupled quantum dot arrays with tunable in-plane coupling and voltage-controlled optical properties of single charge and spin states. △ Less

Submitted 14 April, 2012; originally announced April 2012.

Many-body correlations of electrostatically trapped dipolar excitons

Authors: G. J. Schinner, J. Repp, E. Schubert, A. K. Rai, D. Reuter, A. D. Wieck, A. O. Govorov, A. W. Holleitner, J. P. Kotthaus

Abstract: We study the photoluminescence (PL) of a two-dimensional liquid of oriented dipolar excitons in In_{x}Ga_{1-x}As coupled double quantum wells confined to a microtrap. Generating excitons outside the trap and transferring them at lattice temperatures down to T = 240 mK into the trap we create cold quasi-equilibrium bosonic ensembles of some 1000 excitons with thermal de Broglie wavelengths exceedin… ▽ More We study the photoluminescence (PL) of a two-dimensional liquid of oriented dipolar excitons in In_{x}Ga_{1-x}As coupled double quantum wells confined to a microtrap. Generating excitons outside the trap and transferring them at lattice temperatures down to T = 240 mK into the trap we create cold quasi-equilibrium bosonic ensembles of some 1000 excitons with thermal de Broglie wavelengths exceeding the excitonic separation. With decreasing temperature and increasing density n <= 5*10^10 cm^{-2} we find an increasingly asymmetric PL lineshape with a sharpening blue edge and a broad red tail which we interpret to reflect correlated behavior mediated by dipolar interactions. From the PL intensity I(E) below the PL maximum at E_{0} we extract at T < 5 K a distinct power law I(E) \sim (E_{0}-E)^-|α| with -|α|\sim -0.8 in the range E_{0}-E of 1.5-4 meV, comparable to the dipolar interaction energy. △ Less

Submitted 14 April, 2012; v1 submitted 30 November, 2011; originally announced November 2011.