June 2021 - informatik-vortraege - lists.rwth-aachen.de

Einladung Informatik-Oberseminar Malte Nuhn

by Sekretariat I6

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Freitag, 12. Juli 2019, 10.00 Uhr Ort: Informatikzentrum, E3, Raum 9222 Referent: Dipl.-Inform. Malte Nuhn Thema: Unsupervised Training with Applications in Natural Language Processing// Abstract: The state-of-the-art algorithms for various natural language processing tasks require large amounts of labeled training data. At the same time, obtaining labeled data of high quality is often the most costly step in setting up natural language processing systems.Opposed to this, unlabeled data is much cheaper to obtain and available in larger amounts.Currently, only few training algorithms make use of unlabeled data. In practice, training with only unlabeled data is not performed at all. In this thesis, we study how unlabeled data can be used to train a variety of models used in natural language processing. In particular, we study models applicable to solving substitution ciphers, spelling correction, and machine translation. This thesis lays the groundwork for unsupervised training by presenting and analyzing the corresponding models and unsupervised training problems in a consistent manner.We show that the unsupervised training problem that occurs when breaking one-to-one substitution ciphers is equivalent to the quadratic assignment problem (QAP) if a bigram language model is incorporated and therefore NP-hard. Based on this analysis, we present an effective algorithm for unsupervised training for deterministic substitutions. In the case of English one-to-one substitution ciphers, we show that our novel algorithm achieves results close to human performance, as presented in [Shannon 49]. Also, with this algorithm, we present, to the best of our knowledge, the first automatic decipherment of the second part of the Beale ciphers.Further, for the task of spelling correction, we work out the details of the EM algorithm [Dempster & Laird + 77] and experimentally show that the error rates achieved using purely unsupervised training reach those of supervised training.For handling large vocabularies, we introduce a novel model initialization as well as multiple training procedures that significantly speed up training without hurting the performance of the resulting models significantly.By incorporating an alignment model, we further extend this model such that it can be applied to the task of machine translation. We show that the true lexical and alignment model parameters can be learned without any labeled data: We experimentally show that the corresponding likelihood function attains its maximum for the true model parameters if a sufficient amount of unlabeled data is available. Further, for the problem of spelling correction with symbol substitutions and local swaps, we also show experimentally that the performance achieved with purely unsupervised EM training reaches that of supervised training. Finally, using the methods developed in this thesis, we present results on an unsupervised training task for machine translation with a ten times larger vocabulary than that of tasks investigated in previous work. Es laden ein: die Dozentinnen und Dozenten der Informatik _______________________________________________ -- -- Stephanie Jansen Faculty of Mathematics, Computer Science and Natural Sciences HLTPR - Human Language Technology and Pattern Recognition RWTH Aachen University Ahornstraße 55 D-52074 Aachen Tel. Frau Jansen: +49 241 80-216 06 Tel. Frau Andersen: +49 241 80-216 01 Fax: +49 241 80-22219 sek(a)i6.informatik.rwth-aachen.de www.hltpr.rwth-aachen.de Tel: +49 241 80-216 01/06 Fax: +49 241 80-22219 sek(a)i6.informatik.rwth-aachen.de www.hltpr.rwth-aachen.de

1 year

3
15
0 0

UnRAVeL "Behind the Scenes" Survey Lecture

by Tim Seppelt

Dear all, part of the programme of the research training group UnRAVeL is a series of introductory lectures on the topics of „randomness“ and „uncertainty“ in UnRAVeL’s research thrusts algorithms and complexity, verification, logic and languages, and their application scenarios. Each lecture is delivered by one of the researchers involved in UnRAVeL. The main aim is to provide doctoral researchers as well as master students a broad overview of the subjects of UnRAVeL. This year, 12 UnRAVeL professors will answer the following questions, based on one of their recent scientific results: * How did you get to this result? * How did you come up with certain key ideas? * How did you cope with obstacles on the way? Which ideas you had did not work out? Following these talks, PhD students will give an informal summary of their doctoral studies within UnRAVeL. All interested doctoral researchers and master students are invited to attend the UnRAVeL lecture series 2021 and engage in discussions with researchers and doctoral students. Details information can be found on https://www.unravel.rwth-aachen.de/cms/UnRAVeL/Studium/~pzix/Ringvorlesung-… All events take place on *Thursdays from 16:30 to 18:00 on Zoom* https://rwth.zoom.us/j/96043715437?pwd=U0dRczkyQjRCY21abW13TDNmUHlhUT09 * 15/04/2021 Survey Lecture: Erika Ábrahám: Probabilistic Hyperproperties * 22/04/2021 Jürgen Giesl: Inferring Expected Runtimes of Probabilistic Programs * 29/04/2021 Erich Grädel: Hidden Variables in Quantum Mechanics and Logics of Dependence and Independence * 06/05/2021 Christof Löding: Learning Automata for Infinite Words * 20/05/2021 Martin Grohe: The Logic of Graph Neural Networks * 10/06/2021 Britta Peis: Sensitivity Analysis for Submodular Function Optimization with Applications in Algorithmic Game Theory * 17/06/2021 Nils Nießen: Optimised Maintenance of Railway Infrastructure * 24/06/2021 Gerhard Lakemeyer: Uncertainty in Robotics * 01/07/2021 Joost-Pieter Katoen: The Surprises of Probabilistic Termination * 08/07/2021 Christina Büsing: Robust Minimum Cost Flow Problem Under Consistent Flow Constraints * 15/07/2021 Ringvorlesung: Gerhard Woeginger: Bilevel optimization * 22/07/2021 Ulrike Meyer: Malware Detection We are looking forward to seeing you at the lectures. Best regards, Tim Seppelt for the organisation committee https://www.unravel.rwth-aachen.de/global/show_picture.asp?id=aaaaaaaaaydoc…

1 year, 10 months

1
13
0 0

Einladung: Informatik-Oberseminar Andrea Schnorr

by Andrea Schnorr

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Freitag, 19. Februar 2021, 11.00 Uhr Zoom: https://rwth.zoom.us/j/2452218628 Referent: Andrea Schnorr, M.Sc. LuFG i12 Thema: Feature Tracking for Space-Filling Structures Abstract: Feature-based visualization is a proven strategy to deal with the massive amounts of data emerging from time-dependent simulations: the analysis focuses on meaningful structures, i.e., said features. Feature tracking algorithms aim at automatically finding corresponding objects in successive time steps of these time-dependent data sets in order to assemble the individual objects into spatio-temporal features. Classically, feature-based visualization has focused on sparse structures, i.e. structures which cover only a small portion of the data domain. Given a sufficiently high temporal resolution, existing tracking approaches are able to reliably resolve the correspondence between feature objects of successive time steps. Our research is motivated by our collaborators' work on the statistical analysis of structures that are space-filling by definition: dissipation elements. Space-filling structures partition the entire domain. Our collaborators aim at extending their statistical analysis to a time-dependent setting. Hence, we introduce an efficient approach for general feature tracking which handles both sparse and space-filling data. To this end, we develop a framework for automatic evaluation of tracking approaches, an algorithmic framework for feature tracking, and an efficient implementation of this framework. First, we propose a novel evaluation framework based on algorithmic data generators, which provide synthetic data sets and the corresponding ground truth data. This framework facilitates the structured quantitative analysis of an approach's feature tracking performance and the comparison of different approaches based on the resulting measurements. Second, we introduce a novel approach for tracking both sparse and space-filling features. The correspondence between neighboring time-steps is determined by successively solving two graph optimization problems. In the first phase, one-to-one assignments are resolved by computing a maximum-weight, maximum-cardinality matching on a bi-partite graph. In its second phase, the algorithm detects events by finding a maximum weight independent set in a graph of all possible, potentially conflicting event explanations. Third, we show an optimized version of the second stage of the tracking framework which exploits the model-specific graph structure arising for the tracking problem. The method's effectiveness is demonstrated by a set of case studies including the use of the evaluation framework as well as the analysis of miscellaneous real-world simulation data sets. Es laden ein: die Dozentinnen und Dozenten der Informatik

1 year, 11 months

2
1
0 0

Einladung: Informatik-Oberseminar Herr Peter Marcel de Lange

by Leany Maaßen

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Dienstag, 20. Juli 2021, 10:00-11:00 Uhr Zoom: <https://rwth.zoom.us/j/93538347394?pwd=YXlqZ1VqSE0vRjRnclRtQTY5RVJOZz09> https://rwth.zoom.us/j/93538347394?pwd=YXlqZ1VqSE0vRjRnclRtQTY5RVJOZz09 Meeting-ID: 935 3834 7394 Kenncode: 388699 Referent: Herr Peter Marcel de Lange, M.Sc. Lehrstuhl Informatik 5 Thema: Scaffolding Decentralized Community Information Systems for Lifelong Learning Communities Abstract: With the rise of the Web 2.0, social networking sites and content management systems enabled professional communities to create Web content. But it simultaneously put the communities at the mercy of the platform operators and software providers. Decentralized community information systems bring in a new perspective by offering self-hosting, self-governing and self-developing communities. In this dissertation, we followed a design science approach that provides support for communities to create and host their own, decentralized community information systems. On the one hand, we produced several artifacts to provide possible answers to the question of what properties such an infrastructure needs to fulfill. On the other hand, we transfer the metaphor of educational scaffolding to the domain of service development. We demonstrated and evaluated our open source artifacts on a European scale, with three longitudinal studies conducted within several communities from different areas of technology enhanced learning, such as the European voluntary service, vocational and educational training providers and in higher education mentoring scenarios. Es laden ein: die Dozentinnen und Dozenten der Informatik _______________________________ Leany Maaßen RWTH Aachen University Lehrstuhl Informatik 5, LuFG Informatik 5 Prof. Dr. Stefan Decker, Prof. Dr. Matthias Jarke, Prof. Gerhard Lakemeyer Ph.D. Ahornstrasse 55 D-52074 Aachen Tel: 0241-80-21509 Fax: 0241-80-22321 E-Mail: maassen(a)dbis.rwth-aachen.de

1 year, 11 months

1
0
0 0

Ankündigung: Informatik-Oberseminar

by Thönnessen, David

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Montag, 21. Juni 2021, 13.00 Uhr Ort: Videokonferenz (Zoom-Meeting, Informationen siehe unten) Referent: David Thönnessen, M.Sc. Lehrstuhl Informatik 11 Thema: Hardware-in-the-Loop Testing of Industrial Automation Systems Using PLC Languages Abstract: Testing industrial controllers such as Programmable Logic Controllers (PLCs) poses specific challenges to the test process. Especially in the context of Cyber-Physical Production System (CPPS) the control systems are subject to continuous reconfiguration. Therefore, it is no longer sufficient to test solely before commissioning, but it must be possible to test existing control systems after their reconfiguration with low effort and to put them back into operation. Our approach for this is to develop a test environment that allows an efficient and modular test case specification and can therefore be easily adapted to changing environmental conditions. We have chosen Hardware-in-the-Loop (HiL) simulation as the basis for this test environment since not only the control model or control program is included in the test, but also the control hardware. Our architecture uses slightly extended PLC programming languages to specify test cases. Thus, we avoid the change in methodology that occurs when using dedicated test case specification languages and corresponding test environments. Furthermore, we have provided our concept with the possibility of randomized test case generation, such that a large number of test cases can be generated and tested without a tester having to specify them manually. Our hypothesis is that this will lead to faster and more reliable customizable test cases and thus create the desired agility. The evaluation of our implementation shows that especially developers who are familiar with PLC programming languages can achieve an increase in testing efficiency compared to existing test tools. Furthermore, by randomized testing of Safety Programmable Logic Controllers (Safety PLCs), we show that our test tool can find critical errors in control systems, which would have been found with traditional test methods only to a limited extent. From these results we conclude that the concept presented here is a valuable addition to existing test methods and well-tailored to the challenges of CPPS. Es laden ein: die Dozentinnen und Dozenten der Informatik +********************************************************************** Thema: PhD Defense David Thönnessen Uhrzeit: 21.Juni.2021 01:00 PM Amsterdam, Berlin, Rom, Stockholm, Wien Zoom-Meeting beitreten https://rwth.zoom.us/j/96024814962?pwd=eEU5SmNFTjYyWitJVldVRmQ3TE5pdz09 Meeting-ID: 960 2481 4962 Kenncode: 158094 +**********************************************************************

1 year, 11 months

1
0
0 0

Einladung: Informatik-Kolloquium Prof. Stephan Günnemann, 15. Juni um 14h

by Gerhard Lakemeyer

1 year, 11 months

1
0
0 0

Einladung: Informatik-Oberseminar Herr Sevket Gökay

by Leany Maaßen

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Dienstag, 15. Juni 2021, 14:00-15:00 Uhr Zoom: https://rwth.zoom.us/j/95286197657?pwd=TWtHZHBpa3Q4bmNIZUZXZ21NSmdydz09 Meeting-ID: 952 8619 7657 Kenncode: 048149 Referent: Herr Dipl.-Inform. Sevket Gökay Lehrstuhl Informatik 5 Thema: Scalable Real-Time Ride-Sharing with Meeting Points for Flexible On-Demand Public Transportation Abstract: The landscape of personal transportation is colorful and ever-changing. Ride-sharing, a member of the on-demand transportation family, sits between the private and public transportation categories. It can transport multiple passengers with similar journeys in the same vehicle in a flexible and convenient manner. This work explores real-time ride-sharing in three steps. First, we develop and evaluate a dynamic (i.e. flexible, on-demand) bus-like service as an alternative to the traditional bus service in rural areas with low demand. Simulations conducted in Aachen and in Ulm indicate that both the providers and the customers might benefit from this alternative. Next, we address the computational scalability issue of the service in urban areas with high demand. We present an approach to reduce processing time by employing an improved trip-vehicle fitness heuristic. The evaluation simulates New York City taxi trips in a ride-sharing context, and exhibits significant performance improvement, together with improved customer satisfaction and vehicle costs. Finally, we investigate the prevention of small detours of vehicles, by merging location visits with close proximity into one by introducing small walking paths. The results hint at a significant increase in the number of satisfied trip requests. Es laden ein: die Dozentinnen und Dozenten der Informatik _______________________________ Leany Maaßen RWTH Aachen University Lehrstuhl Informatik 5, LuFG Informatik 5 Prof. Dr. Stefan Decker, Prof. Dr. Matthias Jarke, Prof. Gerhard Lakemeyer Ph.D. Ahornstrasse 55 D-52074 Aachen Tel: 0241-80-21509 Fax: 0241-80-22321 E-Mail: maassen(a)dbis.rwth-aachen.de

2 years

1
0
0 0

Einladung: Informatik-Oberseminar Daniel Wiebking

by Daniel Wiebking

+********************************************************************** * * * Einladung * * * * Informatik-Oberseminar * * * +********************************************************************** Zeit: Montag, 14. Juni 2021, 16:00 Uhr Ort: Zoom Videokonferenz Link: https://rwth.zoom.us/j/94470149252?pwd=ZnYwVXQrbStveVdDNDk1V1BlN1lJUT09 Meeting-ID: 944 7014 9252 Kenncode: 004247 Referent: Daniel Wiebking, M.Sc. Lehrstuhl Informatik 7 Thema: A Decomposition-Compatible Canonization Framework for the Graph Isomorphism Problem Abstract: The Graph Isomorphism Problem is one of a few famous problems in NP that is neither known to be solvable in polynomial time nor to be NP-complete. The problem asks whether two given input graphs are structurally equivalent, which means that both graphs coincide up to a renaming of the vertices. With Babai's celebrated breakthrough (STOC 2016) it was shown that the problem can be decided in quasipolynomial time. One way of solving the Graph Isomorphism Problem is by applying a canonization approach. Graph canonization is the task of transforming a graph into a canonical form, that is, a graph representation that coincides for structurally equivalent graphs. As far as we know, graph canonization might be harder than the Graph Isomorphism Problem. In particular, there are isomorphism tests for various graph classes and objects for which to date no canonization algorithm with the same asymptotic running time is known. In this thesis, we devise a unified canonization framework for graphs, and beyond that for combinatorial objects in general. We use that framework to design new fastest canonization algorithms with an asymptotic running time matching the best known isomorphism tests. Our framework supports the use of decomposition techniques. By combining our framework with new graph-theoretic decompositions, we not only match but even improve the running time of existing isomorphism tests for graphs of bounded treewidth and graphs excluding fixed minors. Our improved algorithms for restricted graph classes come hand in hand with new insights about the automorphism groups. We prove several restrictions on these groups by analyzing them from a purely mathematical point of view. Beyond that, our decomposition-friendly framework has also applications in computational group theory. In particular, we design a new fastest algorithm computing normalizers of permutation groups. Finally, we investigate the connections between isomorphism testing and canonization from a logical perspective. To this end, we provide a new computation model that supports a construction turning isomorphism tests into canonization algorithms. Es laden ein: die Dozentinnen und Dozenten der Informatik

2 years

1
0
0 0

2023

2022

2021

2020

2019

2018

informatik-vortraege June 2021