“Machine learning is possibly the most rapidly evolving field of our era,” says Professor Alexandros Iosifidis

What are your main research interests?

My research focuses on machine learning and computational intelligence, with an emphasis on designing and analysing learning algorithms that enable computers to learn from data in a reliable and meaningful way. I am interested in statistical learning and deep learning methods for structured data, such as time series, graphs and multimodal signals, with applications ranging from computer vision and perception to financial forecasting.

Two central themes in my work are robustness and efficiency: how to develop learning systems that are extremely efficient, and how to ensure they can operate reliably under uncertainty, limited data or changing environments. Ultimately, my goal is to build intelligent systems that can be trusted in real-world settings, rather than only under ideal laboratory conditions.

What makes your research significant?

The significance of my research lies in bridging theoretical machine learning with practical deployment in real-world environments. While many learning methods perform well under controlled conditions, they often struggle when data is imperfect or the environment changes. 

By focusing on robustness, my work helps to close this gap and supports the development of machine learning systems that are not only accurate, but also stable, efficient and adaptable. As machine learning becomes increasingly embedded in critical societal systems, these qualities are essential for ensuring reliability and accountability.

Where do you draw inspiration for your work as a professor?

Much of my inspiration comes from real-world problems that reveal the limitations of current machine learning methods. Seeing where existing approaches fall short motivates me to rethink models and assumptions. 

Teaching and research supervision also play a key role, as explaining complex ideas clearly and brainstorming with early-career researchers often leads to deeper insights and new research directions. 

What would you want to study next and why?

Machine learning – the driving force behind artificial intelligence – is possibly the most rapidly evolving field of our era. This makes research both very challenging and very interesting. In such a dynamic landscape, research interests naturally shift as new challenges and application demands emerge. 

One area that particularly interests me at the moment is the efficient and detailed processing of ultra-high-resolution signals. This is important because the dominant trend in machine learning has been towards increasingly large and computationally expensive models. Although these models can achieve impressive accuracy, they are often impractical when working with ultra-high-resolution data. This raises fundamental questions about representation learning, generalisation under limited resources, and the trade-offs between accuracy, complexity and energy consumption.

What do you do in your free time?

Outside work, I enjoy spending time with my family and engaging in activities that allow me to disconnect from screens and the constant flow of information. I recently started learning Finnish, which I find very challenging, especially given the mental workload of an academic career.