Skip to content
Main menu hidden.

Image: Johan Gunséus

Statistical Learning and Inference for Spatio-Temporal Data

Research group Research in our group includes signal/image analysis and statistical learning and inference for spatio-temporal data.

The research group at the Department of Mathematics and Mathematical Statistics is led by Professor Jun Yu and conducts research in signal/image analysis, statistical learning and inference for spatio-temporal data.

Research expertise

We work on tackling theoretical data science problems and developing statistical learning methods for solving real-life problems, which originate from various application areas, including atmospheric icing, automobile industry, biomedical engineering, climate research, epidemiology, forestry, geochemistry and hydrology, radiation oncology, spatial ecology, sports science, and transportation.

Regarding the statistical learning and inference studied: statistical learning with sparsity, compressive sensing, mathematics of data science, hierarchical spatiotemporal modelling, nonparametric density/intensity estimation and smoothing techniques, statistical inference for hidden Markov models and random fields, summary statistics for point processes, and wavelet theory applied to signal and image analysis.

In terms of data analysis tools: intelligent data sampling using compressive sensing, large-scale environmental data model, multimodal image processing, tree growth models, and general modelling of biological populations in space and time.

Cancer research

Two of the research group's projects are related to cancer research. The purpose of one project is to develop stochastic models and statistical methods that can provide more reliable estimates of the physiological parameters that can be obtained from several different spatiotemporal signals including DCE-MRI and MRS. Analysis methods will be applied, for example, in the characterization of cancerous tumours.

The main purpose of the second project is to develop statistical and computational methodologies for intelligent data sampling and uncertainty analysis of MRI and PET measurements. It aims at contributing to the general understanding of optimised data sampling in bio-imaging and to efficient noise reduction for improved quality of the estimated parametric images in clinical use. 


Collaboration at Umeå University

Umeå School of Sport Sciences
Department of Radiation Sciences: Thomas Asklund, Patrik Brynolfsson, Anders Garpebring, Christer Grönlund, Adam Johansson, Håkan Jonsson, Mikael Karlsson, Anne Larsson Strömvall, Tufve Nyholm, Urban Wiklund
Sports Medicine: Kajsa Gilenstam, Christer Malm, Michael Svensson
Magnus Ekström (Department of Statistics)
Göran Englund (Department of Ecology and Environmental Sciences)
Charlotte Häger (Department of Community Medicine and Rehabilitation)
Shafiq ur Rehman (Department of Applied Physics and Electronics)
Tor Söderström (Department of Pedagogy)
Johan Trygg (Department of Chemistry)
Patrik Eklund (Department of Computing Science)
Mats Johansson (Department of Computing Science)
Lili Jiang (Department of Computing Science)
Maria Eriksson (Department of Plant Physiology)

Main collaborators outside Umeå University:

Anastassia Baxevani (University of Cyprus, Nicosia, Cyprus)
Pavel Grabarnik (Russian Academy of Sciences, Pushchino, Russia)
Juha Heikkinen (The Finnish Forest Research Institute, Finland)
Dinghua Huang (China University of Geosciences, Wuhan, China)
M.N.M. van Lieshout (CWI research centre, Amsterdam, The Netherlands)
Zhengyan Lin (Department of Mathematical Sciences, Zhejiang University, China)
Stefan Löfgren (Department of Aquatic Sciences and Assessment, SLU)
Jorge Mateu (Department of Mathematics, Universitat Jaume I, Castellón, Spain)
Kenneth Nyström (Department of Forest Resource Management, SLU)
Henning Omre (Department of Mathematics, NTNU, Norway)
Nils Östlund (Konftel, Umeå)
Arne Pommerening (Department of Forest Resource Management, SLU)
Bo Ranneby (Centre of Biostochastics, SLU)
Holger Rauhut (Department of Mathematics, RWTH Aachen University, Germany)
Claudia Redenbach (Kaiserslautern University, Germany)
Aila Särkkä (Department of Mathematics Statistics, Chalmers University of Technology)
Mari Myllymäki (LUKE, Finland)
Juha Heikkinen (LUKE, Finland)
Jia Li (Penn State University, USA)
Jonas Örmin (Volvo GTO, Umeå)
Hans Wenngren (Volvo GTO, Umeå)
Kent Sundberg (Volvo GTO, Umeå)
Eric Lindahl (Volvo GTO, Umeå)
Sandra Finér (Volvo GTO, Umeå)
Kendall Rutledge (Novia, Finland)
Niklas Frände (Novia, Finland)
Petri Välisuo (University of Vasa, Finland)
Mohammad Virk (University of Tromsö, Norway)
Javier Martin-Torres (Luleå University of Technology)
Ricardo Fonseca (Luleå University of Technology)
Heli Koivuluoto (Tempere University)

Ida Häggström (Chalmers)


Head of research

Jun Yu


Participating departments and units at Umeå University

Department of Mathematics and Mathematical Statistics

Research area

Cancer, Mathematics, Statistics

Former members

Alex Teterukovsky (Head of Actuary and Head of Product and Price, If Insurance, Sweden)

Yingfu Xie (Time Series and Data Analyst, SCB, Statistics Sweden)

Pia Löthgren (Statistician, Region Skåne, Sweden)

David Bolin (Associate Professor at King Abdullah University of Science and Technology, Saudi Arabia)

Kristi Kuljus (Senior Research Fellow at University of Tartu, Estonia)

Zhiyong Zhou (Professor at Zhejiang University City College, China)

Ottmar Cronie (Associate Professor at Gothenburg University, Sweden)

Fekadu Bayisa (Research Fellow at University of Guelph, Canada)

Jianfeng Wang (Principal Statistician, Kenvue)

Mohammad Ghorbani (Associate Professor at Luleå University of Technology, Sweden)

Robin Rohlén (Postdoctoral Researcher at Lund University, Sweden)

Bild på Klara Leffler som justerar PET-maskinen.
New methods for safer examination of cancer patients

Improved data collection in PET scans increases patient safety and comfort.

Latest update: 2023-02-09