Heikki Mannila

I defended my Ph.D. thesis in 1985 at the University of Helsinki, and after a short while in Tampere was a professor at the University of Helsinki from 1989 to 1999, when I moved to Helsinki University of Technology (now Aalto). After that I had for several years a group at both universities. I have also worked at Technische Universität Wien, Max Planck Institut für Informatik, Microsoft Research in Redmond and Nokia Research. From 2004 to 2008 I was academy professor.

I served as vice president for research and education at Aalto University in 2009-2012, and as the president of the Academy of Finland (Finnish Research Council) from 2012 to 2022.

My research has concentrated on algorithms in data analysis, data mining and machine learning and the interplay of theory and practice, as well as applications in other sciences and in industry. The approach has been to look at application areas, formulate computational concepts, analyse their formal and practical properties, design algorithms, and take the resulting methods to practice. I've done multidisciplinary work together with people from, e.g., environmental sciences, linguistics, and paleontology, as well as on problem stemming from industry.

I'm now working on House of AI, an initiative for multidisplinary research combining other sciences with artificial intelligence and data analysis: House of AI | Aalto University.

Recent papers are The Hadamard decomposition problem (with Aris Gionis and Martino Ciaperoni) and A multilinear Johnson-Lindenstrauss transform (with Petteri Kaski and Antonis Matakos; to appear, SOSA 2025).