The mission of Statistics for Innovation, (sfi)², is to develop core statistical methodologies, strategically necessary to achieve innovation goals in four key sectors: petroleum, finance, marine and health.
The ideal research project at (sfi)² is satisfying four principles:
1. It has a clear and ambitious, very ambitious, innovation aim,
2. ... where we have an advantage with respect to competitors because of our partner’s top know-how, market position, unique data and ideas,
3. ... where we have a scientific competitive edge because of our unique competence in statistics,
4. ... where we put together an internationally leading interdisciplinary research team.
Projects which implement these conditions operate in specific niches, where we can be best and first. Several of our projects are of this type. (sfi)² has important long-term projects, which will continue in the next years and deliver exciting results in terms of innovation and science. Some research projects have already reached operative innovation results, based on novel statistical methods. Validation of these results has started, often within the operation of our partners, on field work. The aim is to check, compare and optimise the innovation we produced. This will take some time, but will eventually show the actual value of our contribution in petroleum, marine, finance and insurance, health and biotechnology.
Statistical methodology is the basis of all research at (sfi)². We maintain a strong basis component in fundamental research, and target it clearly towards current and expected innovation aims. We see already how methods developed in one area are exported in new territories, often for the first time. The activities of (sfi)² might seem spread, but statistical methodology is the common thread and represents a compact core of (sfi)².
Ambitious projects imply the risk of not reaching fully their declared aims. We conclude some research activities with a report or a scientific paper, but without a major impact on innovation; and vice versa, some projects have a significant impact on innovation, are fully operational, but do not call for top novel science.
(sfi)² is a unique construction in international statistics, and there are exciting years to come. We are delivering solid results, at the interface between innovation and science.
Presentation of ongoing projects
- BioInfStat: Apply, adapt and develop new statistical methods to enhance already existing biotechnologies to make them more competitive. Partners: Oslo University Hospital, Biomolex and PubGene. Contact Arnoldo Frigessi or Marit Holden for more information, and if you wish to join this project activity.
- ClimateInsure: Produce new statistical methods and instruments for the evaluation of climate change effects aimed in particular, but not limited to, risk management for the insurance industry. Partners: Gjensidige, Llyod's and London School of Economics. Contact Arnoldo Frigessi or Kjersti Aas for more information, and if you wish to join this project activity.
- ComplexClin: Produce a novel Bayesian statistical approach for design enrichment, to allow for a joint analysis of clinical trials with different complex designs. Increase competence in statistical methods for non-inferiority trials. This project is with Smerud Medical Research. Contact Ingunn Fride Tvete for more information, and if you wish to join this project activity.
- ComplexDepend: Produce scientific results which underpin (sfi)²‘s innovation strategy by providing new statistical methods and computational tools to current and future innovation projects. Progress statistical science in general. This project is with all our partners. Contact Arnoldo Frigessi for more information, and if you wish to join this project activity.
- CustomerLife: Obtain knowledge about the mechanisms that regulate the customer relationship in an insurance company. Introduce new statistical methods and instruments for an innovative understanding and prediction of customer behaviour for the insurance industry. This allows new commercial strategies for pricing, marketing, churn prediction, fraud detection and risk management. The project is with Gjensidige. Contact Ingunn Fride Tvete for more information, and if you wish to join this project activity.
- Elprice: Produce new understanding of and new statistical methods for the electricity market. Contribute to the next generation tools for the management of electricity price risk and production planning. The project is with Norsk Hydro ASA. Contact Anders Løland for more information, and if you wish to join this project activity.
- FindOil: Improve oil exploration efficiency, both by improved data interpretation for single prospects and using joint models for several prospects. The project is with Statoil. Contact Petter Abrahamsen for more information, and if you wish to join this project activity.
- Genestat: Produce statistical instruments to understand molecular mechanisms, on the basis of -omics data, contributing to biological discovery and thus combating diseases. Partners: Oslo University Hospital and PubGene. Contact Arnoldo Frigessi or Marit Holden for more information, and if you wish to join this project activity.
- Infect: Design stochastic models for the spread of infectious diseases between fish farms, and use the model to improve management strategies in the fish farming industry. Stochastic network models for the understanding and management of human infectious diseases. Partners: The Institute of Marine Research and National Veterinary Institute. Contact Magne Aldrin for more information, and if you wish to join this project activity.
- StatMarine: Develop new and powerful statistical methods for better estimation of fish stock abundance with a good evaluation of reliability. Evaluate and aid in rationalizing large and complex sampling programs for research survey and commercial catch data balanced against costs. The project is with The Institute of Marine Research. Contact Magne Aldrin for more information, and if you wish to join this project activity.
- TotalRisk: Renew the tools used for total risk modelling in financial institutions, producing more reliable and useful estimates of the risk. Complex dependency and interactions are key aspects. The project is with DNB. Contact Kjersti Aas for more information, and if you wish to join this project activity.