SAMBA

Statistical Analysis, Machine Learning and Image Analysis - SAMBA

The SAMBA department has comprehensive theoretical and practical knowledge in the fields of statistics, machine learning and image analysis. We are one of Europe's largest and most competent groups within applied statistics and statistical-matematical modelling. We cover a broad spectrum of methods and are a world leader in some of these areas. The appropriate choice of method for the various problems is thus one of our strengths. Many calculations involve uncertainty and the accurate calculation of this quantity is an important speciality.

Research areas

Statistics for climate and environment

Our research within statistics for climate and environment is a mixture of applied, problem-driven projects and fundamental research eligible for publication in scientific journals.

Statistics for finance, insurance and commodity markets

We use statistical methods to quantify all kinds of risk in the banking, insurance and commodity markets. We work a lot with models related to Basel II and Solvency II and have developed a large number of systems related to trading and risk management for various power companies.

Statistics for marine resources and health

We cover a wide range of applications within the areas marine resources and health.

Statistics for technology and industry

We cover a wide range of applications including forecasting, time series analysis, network modelling, optimal delivery, logistics, optimal steering of industry processes, reliability analysis, cost surveys and analysis of very large data sets.

Earth observation

We do research and development of methodologies for analysis of digital remote sensing data from satellites and aircrafts. Application areas include environmental monitoring, climate change monitoring, environmental emergency detection, natural resource mapping and map revision.

Image Analysis and Machine Learning

We have extensive knowledge in the fields of image analysis and machine learning. Object extraction, characterization and recognition from images are central for many of these applications. Another central theme is the analysis of sequences of data, including motion detection and characterization.

Big Insight

This center for research-based-innovation is funded by the Research Council of Norway. Big Insight produces innovative solutions for key challenges facing our consortium of private, public and research partners, by developing original statistical and machine learning methodologies.

Last 5 scientific articles

Hubin, Aliaksandr; Storvik, Geir Olve. Sparse Bayesian Neural Networks: Bridging Model and Parameter Uncertainty through Scalable Variational Inference. Mathematics (ISSN 2227-7390). 12(6) doi: 10.3390/math12060788. 2024.

Lutz, Julia; Roksvåg, Thea Julie Thømt; Dyrrdal, Anita Verpe; Lussana, Cristian; Thorarinsdottir, Thordis Linda. Areal reduction factors from gridded data products. Journal of Hydrology (ISSN 0022-1694). 635 pp 1-12. doi: 10.1016/j.jhydrol.2024.131177. 2024.

Fall, Johanna Jennifer Elisabeth; Gjøsæter, Harald; Tvete, Ingunn Fride; Aldrin, Magne Tommy. Classification of acoustic survey data: A comparison between seven teams of experts. Fisheries Research (ISSN 0165-7836). 274 doi: 10.1016/j.fishres.2024.107005. 2024.

Olsen, Lars Henry Berge; Glad, Ingrid Kristine; Jullum, Martin; Aas, Kjersti. A comparative study of methods for estimating model-agnostic Shapley value explanations. Data mining and knowledge discovery (ISSN 1384-5810). doi: 10.1007/s10618-024-01016-z. 2024. Institutional archive

Redelmeier, Annabelle Alice; Jullum, Martin; Aas, Kjersti; Løland, Anders. MCCE: Monte Carlo sampling of valid and realistic counterfactual explanations for tabular data. Data mining and knowledge discovery (ISSN 1384-5810). doi: 10.1007/s10618-024-01017-y. 2024.