WS 2019/20

DLS-WS2018-19

Alle Vorträge finden jeweils 17:00 Uhr in SR 3325 am Ernst-Abbe-Platz 2 statt!

show Content 20. November 2019 · Hannes Mollenhauer (Helmholtz-Zentrum für Umweltforschung) · Adaptive Information Technologies for Long-term and Event-oriented Environmental Monitoring

Mollenhauer

Adaptive Information Technologies for Long-term and Event-oriented Environmental Monitoring

The constantly accelerating and increasing impact on the environment such as climate and land use change are increasingly complex and represent critically growing challenges of global relevance. Therefore, one of the major challenges of modern environmental research is the establishment of an adequate environmental monitoring infrastructure. For the characterization of ecosystems it is necessary to detect individual processes with suitable monitoring strategies and methods. Due to the natural complexity of all environmental compartments, single point or temporally and spatially fixed measurements are mostly insufficient for an adequate representation. The application of mobile wireless sensor networks for soil and atmosphere sensing offers significant benefits, due to the simple adjustment of the sensor distribution, the sensor types and the sample rate to the local and regional test conditions. Following this approach of long-term and large-scale monitoring of abiotic parameters across scales the entire process, from data collection to data transmission, storage and quality assurance, must be adaptively implemented. To generate such an advanced data management system, a self-organized data transmission system and a generic and transferable semi- as well as full-automated procedures for quality management of abiotic data were derived. In this talk I will discuss our developed adaptive information technologies for long-term and event-oriented environmental monitoring by some field operation examples with an outlook on the application in future research infrastructures.

 

 

 

 

 

 

show Content 04. Dezember 2019 · Tilmann Rabl (Hasso-Plattner-Institut) · Advances in Stream Processing

Rabl

Advances in Stream Processing

The digital transformation is driven by gathering and analyzing ever new sources of data. In many cases, the value of analysis results decreases quickly. The faster decisions can be taken based on data observed, the better. To process data incrementally and in real time, an increasing amount of applications is leveraging stream processing systems. At the same time, new data sources become available and affordable, which means scalable solutions are required.
Current stream processing systems can scale out individual analysis tasks to tens to hundreds of nodes, processing millions of events per second. Today these systems can support the world's largest companies at peak hours, but they are severely limited in the number and dynamicity of analysis tasks making them inflexible and inefficient in setups with many users. In this presentation, we will show novel methods and system architectures that enable highly efficient stream data processing in these dynamic setups.

 

 

 

show Content 22. Januar 2020 · Johannes Blömer (Universität Paderborn) · Soft and Hard Clustering

Bloemer

Soft and Hard Clustering

Clustering is one of the main techniques in data analysis. Usually its goal is to partition a set of objects into disjoint clusters such that objects within a cluster are more similar than objects in different clusters. Clusters then are represented by so-called centroids. In soft clustering objects may belong to several clusters with certain percentages while clusters are again represented by centroids. In some cases and applications centroids from soft clustering are better representatives for the whole data set. Many different algorithms have been designed to solve hard clustering problems approximately optimal. For soft clustering problems much less is known. We present several results for soft clusterings that rely on a general technique to transform solutions to a soft clustering to solutions to hard clustering with very similar properties. We also show that the same techniques can be used to simplify and improve the so-called expectation-maximization heuristic (EM algorithm)to compute mixture models for data sets. This is joint work with Sascha Brauer and Katrin Bujna.

 

 

 

 

 

 

show Content 05. Februar 2020 · Sabine Glesner (Technische Universität Berlin) · Static Analysis and Verification of Embedded and Cyber-Physical Systems

Glesner

Static Analysis and Verification of Embedded and Cyber-Physical Systems

Software has become a central and integral part of many systems and products of the information era. In embedded and cyber-physical systems, software is not an isolated component but instead an integrated part of larger, e.g. technical or mechanical systems. During the last decades, there has been an exponential growth in the size of embedded software, resulting in an increasing need for software engineering methods addressing the special needs of the embedded and cyber-physical domain. 
Especially in cyber-physical systems, security has become a major issue. Systems communicate with each other and with their environments and, hence, need to be open and to have interfaces to the outside. 
Nevertheless, being open also implies that malicious intruders can gain access. A first step to increase security is to understand how data and information in the system can flow.

In this talk, I present research results concerning information flow analysis for systems that are modeled in Matlab/Simulink/Stateflow, which is one of the modeling and programming languages of choice in technical environments. A particular problem arises in these systems as they also show time-dependent behavior. We have modeled this behavior by timed automata and constraint systems. Our results can be directly applied to systems from the automotive domain, which I demonstrate by presenting respective case studies from our experiments. Concludingly, I give an overview over further research topics in my group.