Nihat Ay researches data science fundamentals at HIP

Nihat Ay joined the Hamburg University of Technology (TUHH) in April 2021 as a new professor of Data Science Foundations. He is currently setting up his new institute in the HAMBURG INNOVATION PORT (HIP).

In the future, Prof. Ay wants to use his research to help ensure that individual and also global, sociopolitical decisions can be made based on solid information. Ay sees the role model for this process in us humans: „We are constantly bombarded with raw data from our senses, which we sort and evaluate. We assign a meaning to them and thus understand our surrounding world. This process of understanding is the basis of all action. That’s what I’m trying to transfer to learning, intelligent systems in my research.“

Sharing and discovering knowledge
Artificial, learning systems are already capable of recognizing patterns in data on their own. Often, the systems learn as humans give them solutions based on examples. After a few repetitions, the artificial system should then be able to recognize desired anomalies or errors itself. This technology is used, for example, in medical imaging processes to detect certain diseases. However, Nihat Ay wants to go one step further at the TUHH. His long-term goal is no longer just to pass on existing knowledge to learning systems, but rather to discover knowledge through Data Science.

Embodied intelligence
To this end, the scientist is conducting research in the field of embodied intelligence, among other things. „For example, one can imagine a four-legged robot that learns to walk. Our approach is for the system to develop its very own view of the world with the help of sensory data and use this as the basis for its directional control,“ the mathematician says. This principle could then be applied to more general, complex systems, such as traffic systems for cars, airplanes and ships, or even social systems. „Not only should the robot learn to walk, but also traffic and social systems should, in a more abstract sense, run optimally. My goal is to use the fundamentals of Data Science to provide an answer for how this can best be achieved,“ Ay continued.

Seeing how a complex web of interrelationships emerges from abstract mathematical structures fascinates the TUHH researcher. At the TUHH, Ay also wants to motivate his students to push the boundaries of thinking and to be open-minded about things.

Source: TUHH