Machine learning and intelligent systems are the driving force behind the transformation of artificial intelligence and its subdisciplines. They make it possible to develop key technologies that are used in ever more areas, bringing about tremendous economic and social changes. Identifying the potential that learning systems already offer today and how they can be further developed to overcome future challenges is the central objective of the Konrad Zuse School of Excellence in Learning and Intelligent Systems (ELIZA).
More information about the Zuse School ELIZA and details of how to apply can be found here: www.eliza.school
Graduate school at seven German sites
The graduate school ELIZA is closely linked with the European Laboratory for Learning and Intelligent Systems (ELLIS), Europa’s leading research network for machine learning-focused AI with a total of 36 units and over 300 members. The Zuse School ELIZA offers a wide range of opportunities for a master’s degree or PhD in cooperation with 14 academic institutions in all, spread across seven German ELLIS sites: Berlin, Darmstadt, Freiburg, Heidelberg, Munich, Saarbrücken and Tübingen. In addition, ELIZA collaborates with a number of renowned companies, including Amazon, Facebook, Google, and Volkswagen, as well as start-ups like Aleph Alpha.
Supervision by academic fellows
One central element is a supervision concept that involves students and doctoral candidates being jointly supervised by two fellows at different sites – two academic institutions or one academic institution together with an industry partner. The curriculum encompasses cross-site lectures and classes in which students can participate either in person or via online formats. These include courses on the AI Campus platform funded by the German Federal Ministry of Education and Research (BMBF). Joint research and educational events, such as thematic workshops and summer schools, are on offer as face-to-face formats, as is a programme to enhance interdisciplinary skills.
The 48 academic fellows currently at ELIZA, of which 35 are also fellows of the European Laboratory for Learning and Intelligent Systems (ELLIS), form the core element of the school. Among them are some of the most renowned German AI researchers in disciplines such as machine learning, natural language processing, computer vision and robot learning, and include Marc Toussaint (Berlin), Iryna Gurevych (Darmstadt), Thomas Brox (Freiburg), Lena Maier-Hein (Heidelberg), Daniel Cremers (Munich), Krishna Gummadi (Saarbrücken) and Bernhard Schölkopf (Tübingen). Additionally, the school benefits from its industrial fellows – researchers in industry who co-mentor ELIZA students and doctoral candidates and take part in joint research and training, thereby creating direct links between ELIZA and the AI industry in Germany and beyond.
Funding for underrepresented groups
The Zuse School ELIZA places great value on giving the members of underrepresented groups the chance to acquire a first-class education. To this end, a special scholarship is available for master’s students with an outstanding academic track record. Students who are already enrolled on a master’s programme at one of ELIZA’s partner universities are eligible to apply.
Applications are submitted directly to the Zuse School.
German and international students who have graduated with an excellent bachelor’s degree can apply for the two-year master’s scholarship. Research-oriented master’s scholarships give outstanding master’s students the chance to prepare for a later doctoral degree and to work together with ELIZA research groups even before beginning their final dissertation. In this context, some ELIZA universities also offer consecutive master’s and PhD funding. Bachelor’s graduates or master’s students with an excellent academic track record can apply. In addition, ELIZA provides funding for excellent master’s students from underrepresented groups to enable them to embark on an AI degree programme in Germany. Information about the application process can be found here: www.eliza.school.
If you are interested in the PhD programme, you can apply either via the portal of the PhD programme of the European Laboratory for Learning and Intelligent Systems (ELLIS) or directly to the individual ELIZA fellows. Details of the application process can be found at:
ELIZA research and training activities focus on four main areas:
Foundations of machine learning
The first step is to study and learn the foundations of machine learning – including ML-driven disciplines such as computer vision, natural language processing and robot learning.
On the basis of expertise from the areas of software engineering and computer systems, machine-learning systems will be further developed and scaled up to the kind of major applications needed in industry, for example.
Particular emphasis will be placed on applications in autonomous systems, such as autonomous driving or autonomous production processes in Industry 4.0. In addition, we will study transdisciplinary applications in the natural and life sciences – encompassing everything from medicine to physics.
Resilience and science communication
Another area of focus at the graduate school is on strengthening resilience and competencies in science communication. The goal is to provide master’s students and doctoral candidates with optimal support in difficult phases and to arm them with the knowledge and skills they will need in highly successful phases to communicate their research findings clearly and concisely at scientific conferences and when cooperating with partners in industry.
ELIZA builds upon ELLIS, the European Laboratory for Learning and Intelligent Systems, Europa’s leading academic network for machine learning-focused AI.
Seven German ELLIS units
ELIZA interlinks seven German ELLIS units with their respective academic institutions. In the case of Berlin, these are Technische Universität Berlin, Freie Universität Berlin and the Fraunhofer Heinrich-Hertz-Institut; in the case of Darmstadt it is the Technical University of Darmstadt; in Freiburg it is the University of Freiburg; in Heidelberg it is Heidelberg University, the German Cancer Research Center and the European Molecular Biology Laboratory; in Munich it is the Technical University of Munich; in Saarbrücken it is Saarland University, the Max Planck Institute for Informatics and the Max Planck Institute for Software Systems, and in Tübingen it is the University of Tübingen and the Max Planck Institute for Intelligent Systems.
Partners from industry
Cooperation with industry in both research and training is an important element of the Zuse School ELIZA. To this end, ELIZA maintains a broad set of industry collaborations, ranging from German AI start-ups and large German companies to multinational corporations. Its current partners include: Aignostics, Aleph Alpha, Amazon, BioRN, Bosch / Bosch Center for Artificial Intelligence, Cellzome, Facebook, Google, Intel, Merantix, Mercedes-Benz, Siemens, Swift, Toyota Motor Europa, Valeo.AI and Volkswagen.
Professor Stefan Roth, coordinator of the Zuse School ELIZA and head of the ELLIS Unit at TU Darmstadt, explains what master’s and PhD students from around the world can expect at the Konrad Zuse School of Excellence in Learning and Intelligent Systems (ELIZA).
Professor Roth, what sets ELIZA apart as an excellent Zuse School for budding AI experts?
Our 14 cooperation partners are spread all over Germany, a constellation that is certainly a unique feature. Another important point is that we bring together Germany’s leading experts in the field of machine learning. The school resulted from the European initiative ELLIS, an alliance of European researchers that aims to raise the profile of AI in Europe and to develop sustainable structures. As seven of the German ELLIS sites are represented at ELIZA, we regard the Zuse School as Germany’s contribution to this network. Our strength is our ability to pool the expertise of these research partners in the area of machine learning, which of course hugely benefits the master’s and PhD students at ELIZA.
What do the cooperation partners hope the programme will achieve?
Worldwide, countries are investing massively in machine-oriented applications in AI. One of our goals therefore is not only to maintain but also to build on Germany’s strengths in those areas of AI that are driven by machine learning and in fields such as robotics. A second goal concerns the international competition for the best talents. To be able to compete, we need to be able to make attractive offers. However, there are various hurdles in Germany that make this more difficult. For example, students in Germany typically begin a PhD only once they have completed their master’s, which is not the case in other countries such as the US or Canada. Consequently, we lose out on international talents who only have a bachelor’s degree – they end up at the top institutions in North America. We want to counter this by offering not only attractive financial conditions but also a great support, supervision and research programme.
What makes this supervision programme so special?
Providing support and supervision across our various sites is a central component of our programme. All master’s and PhD students that we teach at ELIZA are supervised by two fellows, each at different sites. This is helpful for the young researchers, as it allows them to pursue far more intensive discussions than are normally possible with the conventional one-person supervision system. This form of supervision also has advantages for the universities themselves. Having researchers with complementary perspectives who are socialised and work in different academic contexts always advances the institution’s own research. We make this exchange possible at a very early stage.
We work closely together with very well-known German and international companies – from start-ups in the area of AI to automotive giants or globally active internet firms.
What role is played by the interlinking of science and business within the framework of ELIZA?
We work closely together with very well-known German and international companies – from start-ups in the area of AI to automotive giants or globally active internet firms. If one looks at how cooperation with industry happens in other countries, it is clear that it is driven by product development to a much lesser extent than in Germany. In Silicon Valley, collaboration with industry takes place at the research level. Joint publications are produced, and there is direct and reciprocal cross-pollination between industrial and university research at the level of application-oriented basic research. In Germany, by contrast, cooperation very quickly drifts into the area of product development, which in my view ignores some important aspects. One is the question of recruitment: many AI experts want to continue writing papers rather than devoting their work solely to software development. We are concerned to see that many of our top talents are going straight from the university labs to major non-European internet companies. One reason for this is that German firms are less open to research. Through ELIZA, we hope to play our part in ensuring that relations between research institutions on the one hand and companies on the other change in this respect.
TU Darmstadt was awarded the contract in July 2022. What are the next steps?
One challenge for our Zuse School will be to interlink the 14 sites in seven different cities. One essential part of the programme will therefore be events that bring PhD and master’s students together at ELIZA, alternating between the different locations. Our first job in Darmstadt will be to set up a strong team to oversee the administrative side of these processes. And of course to reach out to international AI talents it is crucial to portray ourselves in a good light, which is why a science manager will be expressly assigned to this task and will develop a recruiting strategy. As far as international visibility is concerned, the DAAD is naturally especially important for ELIZA thanks to its global presence and experience, so it will accompany us throughout these initial steps, providing ideas and advice.
To what extent does Darmstadt as a research location have an international orientation?
Darmstadt is a high-tech site with a much more international orientation than one might imagine for a city of its size – as it is home to the ESA’s Satellite Control Center, one could say that Darmstadt is Europe’s Houston.
ELLIS also enjoys a high international profile, its PhD programme having recently attracted nearly 2,000 applications from around the world. This trend is also reflected at the university. In my own department I have only one German PhD student at present, all the others have an international background.
The German state of Hesse is pushing forward the transfer of AI into business and society. How closely do you cooperate with the Hessian Center for Artificial Intelligence?
hessian.AI is an interstate institution that connects all higher education institutions and is intended to shape and enhance the AI landscape in Hesse as a whole. Even though its goals are rather different, there are lots of areas of overlap with our objectives. We also have close personal relationships with the people there. Almost all ELIZA fellows were founding members of hessian.AI, and its two directors – Kristian Kersting and Mira Mezini – are likewise fellows at ELIZA. hessian.AI is well positioned to support start-ups, which benefits us a great deal. Our scholarship holders profit from these close links to the AI industry as they can acquire practical experience there.
Interview: Gunda Achterhold