I am advertising a postdoc position at the intersection of topological data analysis, natural language processing and machine learning.
duration: 3 years starting date: 1 October 2021 location: Heinrich Heine University Düsseldorf, Germany
Reflecting its interdisciplinary nature, the position is associated both with my group (Topology and Geometry) and with the research group of Prof. Milica Gasic (Dialogue Systems and Machine Learning). Prior experience in one or more of the named areas would be beneficial, but we are willing to consider outstanding candidates from other areas of mathematics.
The position will be paid at the TV-L E13 scale (see https://www.academics.com/guide/salary-researchers-germany for more information). It comes with moderate teaching duties (2 times 90 minutes per week during teaching periods), which could be arranged to be of direct relevance for the intended research. Working knowledge of German would be beneficial but is not essential.
We (Henry, Antonio, Hanka, Teresa) are organizing open poster sessions via Zoom, with the intention of giving a platform to younger members of our community to showcase their work, as well as a place for everyone to present their most recent research results. Topics of interest will span the full range of applied and computational topology.
The sessions will be two hours long, and our first poster session will take place on Friday, October 8th, 2021 at 11am Eastern time. To submit a title and abstract, by September 24 please fill out the registration form at https://forms.gle/5EocHz2zSwGtLPFF6
Zoom coordinates and the website with the program will be sent out to the AATRN email list several days before each poster session.
Best wishes, Henry Adams, Hana Dal Poz Kouřimská, Teresa Heiss, Antonio Rieser
***** Call for Papers: Workshop on Applications of Topological Data Analysis to Big Data at IEEE BigData 2021 *****
Topological Data Analysis (TDA) is a growing area of research that focuses on studying the “shape” of data. Tools (such as persistent homology) from TDA have been successful in many application areas, including signal processing, computer vision, dynamical systems, and geospatial systems. This workshop will be a venue for researchers and practitioners of TDA to present innovative, state-of-the-art topological methods and novel applications of topological tools to big data.In this workshop, we invite submissions on recent applications of topological data analysis, with a focus on larger datasets. Topics include but are not limited to:
TDA applications to graphs, hypergraphs and networks
TDA applications in natural language processing
TDA applications to image and video analysis
TDA applications in dynamical systems and signal processing
Topological approaches to spatial and social systems
Topological approaches to machine learning
September 30, 2021: Due date for full workshop paper submission (11:59pm AoE)
November 1, 2021: Notification of paper acceptance to authors
November 20, 2021: Camera-ready version of accepted papers due
December 15–18, 2021: Workshop (one day in this date range)
This workshop will be held remotely. Additional information can be found at https://www.egr.msu.edu/~tymochko/TDA_IEEEBigData2021.html.Thank you, Tegan Emerson (Pacific Northwest National Lab) Mason A. Porter (University of California, Los Angeles) Sarah Tymochko (Michigan State University & Pacific Northwest National Lab) Wlodek Zadrozny (University of North Carolina at Charlotte)
If you are interested in participating, please register here before the 9th of August.
We look forward to see you there,
With best regards,
Nicolas Berkouk & François Petit
The theory of one-parameter persistence was developed in the early 2000’s as an attempt to define topological descriptors of datasets which are robust to noise – the so-called barcodes. Since then, it has led to important advances in diverse areas: material sciences, time series analysis, neuroscience, and neural networks, just to name a few. Although well understood, one-parameter persistence faces several limitations such as sensitivity to outliers or to the user’s choice of the function filtering the data. To overcome these shortcomings, the persistence community has introduced a generalization of the above construction: multi-parameter persistence. In this context, the algebraic theory of multi-parameter persistence becomes more complex, and the need for more refined techniques originating from algebraic topology and algebraic geometry more evident. In this symposium, the speakers will present their advances in the field of topological data analysis with a sheaf theoretical or homotopical perspective. They will emphasize the benefits of taking a more abstract point of view on theoretical as well as on applied problems and explain the challenges of computability raised by these approaches.
Organizers: Nicolas Berkouk & François Petit
Program (Eastern Time)
9:00-9:25 Sheaves As Data Structures, abstract , Robert W. Ghrist, University of Pennsylvania, U.S.
9:30-9:55 Computational Topology in Intersection Theory, abstract, Vidit Nanda, University of Oxford, United Kingdom; Martin Helmer, University of California, Berkeley, U.S.
10:00-10:25 Distances on Sheaves, abstract, François Petit, Sorbonne Universités, France; François Petit, Université de Paris, France
10:30-10:55 The Amplitude of An Abelian Category: Measures in Persistence Theory, abstract, Nina Otter, University of California, Los Angeles, U.S.
2:40-3:05 Multi Parameter Persistence on Crossroads of Homotopy Theory and Statistics, abstract, Wojciech Chacholski, KTH Royal Institute of Technology, Sweden
3:10-3:35 Homotopy Invariant Notions of Interleaving and Applications, abstract, Luis Scoccola, Michigan State University, U.S.
3:40-4:05 Persistent Homotopical Algebra, abstract, Grégory Ginot, Université Paris 13, France
4:10-4:35 The Truncated Interleaving Distance for Reeb Graphs, abstract, Elizabeth Munch, Michigan State University, U.S.
Invited Speakers: Hélène Barcelo (Arizona State) Saugata Basu (Purdue) Ulrich Bauer (Technical University of Munich) Andrew Blumberg (Columbia) Peter Bubenik (Florida) Gunnar Carlsson (Stanford) Herbert Edelsbrunner (ISTA) Alexander Grigor’yan (Bielefeld) Facundo Memoli (Ohio State) Elizabeth Munch (Michigan State) Nina Otter (UCLA) Leonid Polterovich (Tel Aviv) Eric Sedgewick (De Paul) Vin de Silva (Pomona College) Katharine Turner (Australian National University)
In addition there will be contributed talks. A call for submission of abstracts for these talks and posters will follow.
Scientific Committee: Jacek Brodzki (University of Southampton) Frédéric Chazal (INRIA) Kathryn Hess (EPFL Lausanne) Brittany Fasy (Montana State University) Robert Ghrist (University of Pennsylvania) Matt Kahle (Ohio State University) Claudia Landi (Università di Modena e Reggio Emilia) Primoz Skraba (Queen Mary, University of London) Schmuel Weinberger (University of Chicago)
We are pleased to report that Conference Proceedings of ATMCS10 will be published in conjunction with the Journal of Applied and Computation Topology (APCT). All those contributing to the conference will be invited to submit research and survey papers.
The conference is supported by the Centre for Topological Data Analysis. Limited financial help will be available.
We look forward to welcoming you next year in Oxford!
Heather Harrington, Ulrike Tillmann and Vidit Nanda
Call for papers Smart Tools and Applications in Graphics 26-29 October 2021
STAG aims to offer a forum for discussing novel ideas and results in Computer Graphics and Visual Computing. Papers addressing both theoretical and application-oriented aspects of research are welcome. STAG also encourages dialogue and cross-fertilization between different fields, including Computer Graphics, Computer Vision, Artificial Intelligence, and Topological Data Analysis.
Computational Geometry, Computational Topology, Topological Data Analysis for Visual Computing are included in the list of topics of STAG 2021.
We’re pleased to announce the VII Mexican Workshop in Geometric and Topological Data Analysis, which will take place online September 22nd-29th, 2021. Registration is free but required, and we are accepting proposals for contributed talks and posters until July 12th. The mini-courses for this year’s workshop are:
Computational Homology, Dynamics, and Data Tomasz Kaczynski, Université de Sherbrooke
What we offer: We offer a postdoc position within the Department of Mathematical Sciences as part of the project “Deciphering Nanoporosity of Amorphous Materials using Topological Data Analysis”. The selected candidate will be conducting research supervised by Lisbeth Fajstrup and Christophe A.N. Biscio and in collaboration with the Department of Chemistry and Bioscience to find new topological descriptors of porosity. The selected candidate will visit Lawrence Berkeley National Laboratory, USA for 4 months to collaborate with Staff Scientist Dmitriy Morozov.
What we request: The selected candidate will hold a PhD in Mathematics or a related field and will have promising research results within Topological Data Analysis, Computational Topology or related subjects. The project to which the postdoc will contribute involves developing the theory of non-monotone and multidimensional persistence to describe variations of the free volume and in particular tunnels in amorphous materials and implementation of such insight as a tool. This will require a strong background in Topological Data Analysis either theoretical or computational.
You may obtain further professional information from Associate Professor Christophe Biscio, phone: +45 9940 8925, e-mail: firstname.lastname@example.org or Head of Department Søren Højsgaard, phone: +45 9940 8801, e-mail: email@example.com
Appointment as Postdoc presupposes scientific qualifications at PhD–level or similar scientific qualifications. The research potential of each applicant will be emphasized in the overall assessment. Appointment as a Postdoc cannot exceed a period of four years in total at Aalborg University. https://www.stillinger.aau.dk/vis-stilling/?vacancy=1153647
Organizer(s): Hengrui Luo, Lawrence Berkeley National Laboratory
Chair(s): Chul Moon, Southern Methodist University
3:35 PM EDT 317318 Characterizing heterogeous information in persistent homology with applications to molecular structure modeling
Speaker. Zixuan Cang, University of California, Irvine
Abstract. Persistent homology is a powerful tool for characterizing the topology of a dataset at various geometric scales. However, in addition to geometric information, there can be a wide variety of nongeometric information, for example, there are element types and atomic charges in addition to the atomic coordinates in molecular structures. To characterize such datasets, we propose an enriched persistence barcode approach that retains the non-geometric information in the traditional persistence barcode. The enriched barcode is constructed by finding the smoothest representative cocycles determined by combinatorial Laplacian for each persistence pair. We show that when combined with machine learning methods, this enriched barcode approach achieves state-of-the-art performance in an important real-world problem, the prediction of protein-ligand binding affinity based on molecular structures.
3:55 PM EDT 317284 Gromov-Wasserstein learning in a Riemannian framework
Speaker. Samir Chowdhury, Stanford University
Abstract. Geometric and topological data analysis methods are increasingly being used to derive insights from data arising in the empirical sciences. We start with a use case where such techniques are applied to human neuroimaging data to obtain graphs which can then yield insights connecting neurobiology to human task performance. Reproducing such insights across populations requires statistical learning techniques such as averaging and PCA across graphs without known node correspondences. We formulate this problem using the Gromov-Wasserstein (GW) distance and present a recently-developed Riemannian framework for GW-averaging and tangent PCA. Beyond graph adjacency matrices, this framework permits consuming derived network representations such as distance or kernel matrices, and such choices lead to additional structure on the GW problem that can be exploited for theoretical and computational advantages. We show how replacing the adjacency matrix representation with a spectral representation leads to theoretical guarantees allowing efficient use of the Riemannian framework as well as state of the art accuracy and runtime in graph learning tasks such as matching and partitioning.
4:15 PM EDT 317312 Density estimation and modeling on symmetric spaces
Speaker. Didong Li, Princeton University
Abstract. In many applications, data and/or parameters are supported on non-Euclidean spaces. It is important to take into account the geometric structure of manifolds in statistical analysis to avoid misleading results. In this talk, we consider a very broad class of manifolds: non-compact Riemannian symmetric spaces. For this class, we provide statistical models on the tangent space, push these models forward onto the manifold, and easily calculate induced distributions by Jacobians. To illustrate the statistical utility of this theoretical result, we provide a general method to construct distributions on symmetric spaces, including the log-Gaussian distribution as an analogue of the multivariate Gaussian distribution in Euclidean space. With these new kernels on symmetric spaces, any existing density estimation approach designed for Euclidean spaces can be applied, and pushed forward to the manifold with an easy-to-calculate adjustment. We provide theorems showing that the induced density estimators on the manifold inherit the statistical optimality properties of the parent Euclidean density estimator; this holds for both frequentist and Bayesian nonparametric methods.
4:35 PM EDT 317251 Convergence of persistence diagram in the subcritical regime
Speaker. Takashi Owada, Purdue University
4:55 PM EDT 317225 Combining geometric and topological information for boundary estimation
Speaker. Justin Strait, University of Georgia
Abstract. We propose a method which jointly incorporates geometric and topological information to estimate object boundaries in images, through use of a topological clustering-based method to assist initialization of the Bayesian active contour model. Active contour methods combine pixel clustering, boundary smoothness, and prior shape information to estimate object boundaries. These methods are known to be extremely sensitive to algorithm initialization, relying on the user to provide a reasonable initial boundary. This task is difficult for images featuring objects with complex topological structures, such as holes or multiple connected components. Our proposed method provides an interpretable, smart initialization in these settings, freeing up the user from potential pitfalls. We provide a detailed simulation study, and then demonstrate our method on artificial image datasets from computer vision, as well as real-world applications to skin lesion and neural cellular images, for which multiple topological features can be identified.
5:00 PM EDT Discussion and Floor-time
This event is a subsequent event from last year’s https://appliedtopology.org/tda-at-jsm/