Statistical analysis and machine learning

We develop solutions for our clients by extracting insights from all types of data. We have theoretical and applied expertise in statistical modelling and machine learning, including natural language processing. In close collaboration with our clients and collaboration partners, we develop models, carry out analyses and implement operational systems.

A critical component of our work is to assist our clients in finding the next best piece of data to further innovation. Our researchers also make valuable theoretical contributions to new methodologies and extensions of established methodologies.  
Statistical modelling and machine learning are at the heart of artificial intelligence. Accurate forecasting, uncertainty quantification, risk assessment and classification require an understanding of both classical statistical methods and modern machine learning methods. Our experienced researchers master the art of choosing the right methodology for a given practical problem. This is fundamental for understanding possibilities and limitations, for understanding the usefulness of available data and for implementing user-friendly solutions. 
In SAMBA, we work with applications in a wide range of areas including finance and insurance, technology, industry, medicine, health, marine, climate and environment. We develop, for example,  

  • methods for detection of anomalies in time series 
  • methods for explainable artificial intelligence 
  • risk models for banks and insurance companies 
  • electricity price forecasts 
  • seasonal climate forecasts for temperature and precipitation 
  • models for the spread of salmon lice between fish farms 
  • language models for anonymising text documents 

Our researchers have a strong background in statistical analysis, mathematical modelling and machine learning, and we work closely with experts from our customers and partners. Our projects are financed by private companies, public agencies, the Research Council of Norway and the European Commission.