Photo of Fredrik Lindsten

Fredrik Lindsten

Professor, Head of Division

I develop machine learning models that can extract information from complex data and make statistical predictions, with a broad spectrum of applications from weather forecasting to materials science.

Presentation

Machine learning is about creating computer programs that can learn to solve complex problems from data. The technology has become part of our daily lives through services like ChatGPT, but it can also be used within various scientific fields.

In my research, I develop machine learning models that can extract information from complex data and make statistical predictions. One example is weather forecasting: a model is trained on historical weather data and can then generate hypothetical future weather states. The forecasts can not only be made accurate, but the model can also indicate how confident it is, which enables, for example, rational action plans when there's a risk of extreme weather. Another research direction is materials science, where machine learning models can be trained on large databases of known materials and then generate new, hypothetical combinations of atoms. In this way, we can more quickly find materials with unique properties, which ultimately can contribute to everything from energy storage to new methods for water purification.

For more information about my background and my research, please visit my external page.

Publications

2026

Louis Ohl, Martin Dahlqvist, Jonas Björk, Sanjay Chakraborty, Fredrik Heintz, Johanna Rosén, Fredrik Lindsten (2026) Prediction of MAX-to-Mxene Convertibility Using Constrained Clustering The Journal of Physical Chemistry C (Article in journal) https://dx.doi.org/10.1021/acs.jpcc.6c02801
Martin Andrae, Erik Wikingsson, So Takao, Tomas Landelius, Fredrik Lindsten (2026) DAISI: Data Assimilation with Inverse Sampling using Stochastic Interpolants (Conference paper)

2025

Martin Andrae, Tomas Landelius, Joel Oskarsson, Fredrik Lindsten (2025) Continuous Ensemble Weather Forecasting with Diffusion models The Thirteenth International Conference on Learning Representations (ICLR 2025), p. 26392-26416 (Conference paper)
Arnaud Doucet, Victor Elvira, Fredrik Lindsten, Joaquin Miguez (2025) Preface special issue onsequential monte carlo methods FOUNDATIONS OF DATA SCIENCE, Vol. 7 (Article in journal) https://dx.doi.org/10.3934/fods.2025019
Filip Ekström Kelvinius, Oskar Andersson, Abhijith S. Parackal, Dong Qian, Rickard Armiento, Fredrik Lindsten (2025) WyckoffDiff- A Generative Diffusion Model for Crystal Symmetry Proceedings of the 42nd International Conference on Machine Learning, p. 15130-15147 (Conference paper)

Research

News

About the Division

Staff at STIMA

About the Department