Data Understanding Lab is a joint spin-off of the University of Siegen and the University of Economics in Katowice. We conceptualise, develop and apply machine learning algorithms for automatic recognition and discovery of regularities in big data collections, mainly sensory data. Rather than simple processing, our software aims to semantically interpret data towards humanlike context description and understanding. We closely cooperate with the Research Group for Pattern Recognition from the University of Siegen and transfer our joint scientific outcomes into the application level in a number of collaborative projects.
The innovation process at Data Understanding Lab starts at the scientific level. We run our own Research Lab employing and supervising PhD Candidates and Postdocs in the area of machine learning. Their most innovative scientific results stimulate our Development Lab which subsequently transforms them into professionally engineered software packages. This technological basis is finally exploited by our Product Lab, where concrete customer applications are addressed by specific software solutions. At all three levels of the innovation process, Data Understanding Lab collaborates with internationally recognised organisations.
Research Lab: In our Research Lab, Master Students, PhD Candidates and Postdocs conceptualise, realise, investigate and scientifically evaluate new methods for adaptive machine learning in close collaboration with scientists of the Research Group for Pattern Recognition from the University of Siegen. Our vision is to propose approaches, which on the one hand remain generic and methodologically universal, but on the other hand make use of domain-specific knowledge and adapt to concrete application domains. The aim is to conceptualise generic pattern recognition algorithms that significantly benefit from the context knowledge of the application. Moreover, our methods are supposed to adaptively optimise the data representation in runtime regarding its discriminative properties. Sensory data fusion and analysis for behavioural biometry, context-based multimedia event detection, analysis and interpretation of medical data as well as knowledge modelling for semantic context description belong to the most important scientific fields of our Research Lab.
Development Lab: The most innovative scientific findings discovered by the Research Lab are taken over by our Development Lab which transforms them into professionally engineered software packages. Here, generic software architectures for adaptive machine learning methods are proposed, developed and tested. At this level of our innovation process we provide abstract software interfaces which are later concretised in our Product Lab in order to address particular customer applications. The software packages developed at this stage remain generic, highly modularised and multidimentionally parametrised, so that they can be applied to different customer problems once their parameters have been optimised towards a concrete application by our Product Lab team. The Development Lab provides software classes, functions and interfaces rather than specific end-user solutions.
Product Lab: In our Product Lab, the abstract software interfaces provided by the Development Lab are extended and transferred into concrete solutions addressing specific customer applications. Our methods and techniques conceptualised and developed at earlier stages of the innovation process are highly generic and multidimentionally parametrised. Fixing the values of the parameters in context of a particular application domain (parameter optimisation) is one of the most important objectives of our Product Lab team. This is necessary for releasing end-user products and may be performed using implicit and explicit feedback as well as adaptive software schemes. Evolutionary algorithms are often applied to achieve this context-awareness. Apart from parameter optimisation and functionality implementation, accessibility and usability aspects of our software solutions come into consideration at this stage. For this, we continuously perform comprehensive end-user tests and assessments. Our products address applications in the context of current critical societal challenges including demographic changes, diseases of modern civilisation, global energy transition, transportation and mobility, etc. Ambient assisted living, smart housing, intelligent transportation as well as mental disorder interpretation make up the most important application domains we consider for the time being.
The career development programme at Data Understanding Lab complies with the innovation process depicted in the figure above. Scientific supervision, technological training as well as international collaboration and exchange opportunities continuously extend your knowledge and skills in the field of applied machine learning. Thereby, the spectrum of your career opportunities (both academic and industrial) within and outside of Data Understanding Lab enlarges significantly over time. Appropriate salary alongside amazing career development opportunities keep you highly motivated which turns out to be remarkably beneficial to our company.
Ideally, you start working with us as a Master Student with strong skills in computer programming and mathematics (earlier or later starts are also possible). Some experience in machine learning, pattern recognition, or data analytics would be helpful, but is not obligatory. First, we define an innovative topic for your master thesis in one of our current machine learning projects together that you then conduct under professional scientific supervision as a member of our Research Lab. Normally, this is finalised within one year. Subsequently, based on the success of the master thesis, you continue to work with us towards your doctorate. In your PhD project, you are fully dedicated to one of our research areas being supervised by a machine learning professor. You finalise your doctorate within four years with a PhD examination taking place at one of the universities in Poland or in Germany. During the PhD project you get opportunities for international internships in globally recognised machine learning institutes in Europe.
After your PhD, you need to choose between the academic and the industrial career path. In the former case, a postdoc phase of at least two years in an academic institute abroad would be highly recommendable. You are very likely to be able to find an appropriate opportunity among our multiple international academic collaboration partners. In the latter case, depending on the project situation at Data Understanding Lab, you may get the opportunity to join our Development Lab or the Product Lab. In the Development Lab you would extend your knowledge and skills in professional software engineering. In the Product Lab you would gain experience in software maintenance taking into consideration accessibility and usability aspects as well as parameter optimisation.
If you are interested in joining the Data Understanding Lab team, please send your application (single PDF) to firstname.lastname@example.org.