+**********************************************************************
*
*
* 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
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…
+**********************************************************************
*
*
* 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
**********************************************************************
*
*
* Einladung
*
*
*
* Informatik-Oberseminar
*
*
*
+**********************************************************************
Zeit: Dienstag, 15. Dezember 2020, 10:00 Uhr
Zoom:
https://rwth.zoom.us/j/99233095930?pwd=dHhTV253V1ZYUzRtSkk1L3A1REZVUT09
Meeting-ID: 992 3309 5930
Kenncode: 626162
Referent: Philipp Weidel, Dipl. Inform.
Thema: Learning and decision making in closed loop simulations of
plastic spiking neural networks
Abstract:
To understand how animals and humans learn, form memories and make
decisions is a highly relevant goal both for neuroscience and for fields
that take some inspiration from neuroscience, such as machine learning
and artificial intelligence. Many models of learning and decision making
were developed in the fields of machine learning, artificial
intelligence, and computational neuroscience. Although these models aim
to describe similar mechanisms, they do not all pursue the same goal.
These models can be differentiated between models aiming to reach
optimal performance on a specific task (or set of tasks) and models
trying to explain how animals and humans learn. Some models of the first
class use biologically inspired methods (such as deep learning) but are
usually not biologically realistic and are therefore not well suited to
explain the function of the brain. Models in the second class focus on
being biologically plausible to explain how the brain works, but often
demonstrate their capability on too simplistic tasks and yield low
performance on well-known tasks from machine learning. This work aims to
close the gap between these two types of models.
In the first part of this talk, tools are described that allow the
combination of biologically plausible neural network models together
with powerful toolkits known from machine learning and robotics. To this
end, MUSIC, the middleware for spiking neural network simulators such as
NEST and NEURON is interfaced with ROS, a middleware for robotic
hardware and simulators such as Gazebo. This toolchain is extended with
interfaces to reinforcement learning toolkits such as the OpenAI Gym.
The second part addresses the question of how the brain can represent
its environment in the neural substrate of the cortex and how a
realistic model of reinforcement learning can make use of these
representations. To this end, a spiking neural network model of
unsupervised learning is presented which is able to learn its input
projections such that it can detect and represent repeating patterns. By
using an actor-critic reinforcement learning architecture driven by a
realistic dopamine modulated plasticity rule the model can make use of
the representations and learn a range tasks.
Es laden ein: die Dozentinnen und Dozenten der Informatik
--
Prof. Dr. Abigail Morrison
IAS-6 / INM-6 / SimLab Neuroscience
Jülich Research Center
&
Computer Science 3 - Software Engineering
RWTH Aachen
http://www.fz-juelich.de/inm/inm-6http://www.fz-juelich.de/ias/jsc/slnshttp://www.se-rwth.de
Office: +49 2461 61-9805
Fax # : +49 2461 61-9460
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------
Forschungszentrum Juelich GmbH
52425 Juelich
Sitz der Gesellschaft: Juelich
Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Volker Rieke
Geschaeftsfuehrung: Prof. Dr.-Ing. Wolfgang Marquardt (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt
------------------------------------------------------------------------------------------------
------------------------------------------------------------------------------------------------