Professor II
Geir Olve Storvik
- Avdeling Statistisk modellering og maskinlæring
- Telefonnummer +47 22 85 58 94
Publikasjoner
- 140 publikasjoner funnet
Palomares, Alfonso Diz-Lois og Storvik, Geir Olve. (2026).
Parameter estimation in Conditional Sequential Monte Carlo algorithms through Particle Learning. Department of Mathematics and Statistics of the University of Helsinki
NVA
Vitenskapelig foredrag
Vis sammendrag
In this work, we explore particle learning strategies for the joint estimation of static parameters and latent states within conditional sequential Monte Carlo (CSMC) algorithms. Building on this idea, we propose the p(parameter)-CSMC algorithm, which incorporates both parameter learning and ancestor sampling, leading to much better mixing properties compared to Gibbs sampling in settings where strong internal correlations may challenge effective exploration. We include an application to the estimation of weights in a branching process model against synthetic data and show that, in this setting, performance is dramatically enhanced, with substantially faster mixing and markedly reduced autocorrelation compared with standard particle Gibbs implementations.
Storvik, Geir Olve. (2025).
Maskinlæring versus statistisk modellering. Den Norske Aktuarforening
NVA
Faglig foredrag
Storvik, Geir Olve; Engebretsen, Solveig; Blasio, Birgitte Freiesleben De og Frigessi, Arnoldo. (2025).
Flaws in the Article “Nearly Instantaneous Time-Varying Reproduction Number for Contagious Diseases—a Direct Approach Based on Nonlinear Regression".
Engebretsen, Solveig; Thorarinsdottir, Thordis Linda; Palomares, Alfonso Diz-Lois; Storvik, Geir Olve; Frigessi, Arnoldo og Blasio, Birgitte Freiesleben De. (2025).
Contribution to the Discussion of 'Some statistical aspects of the Covid-19 response' by Wood et al.
Hubin, Aliaksandr; Storvik, Geir Olve; Frommlet, Florian og Lachmann, Jon. (2024).
A subsampling MCMC approach for Bayesian model selection and model averaging. Universitat Pompeu Fabra
NVA
Vitenskapelig foredrag
Hubin, Aliaksandr og Storvik, Geir Olve. (2024).
Sparse Bayesian Neural Networks: Bridging Model and Parameter Uncertainty through Scalable Variational Inference.
Vis sammendrag
Bayesian neural networks (BNNs) have recently regained a significant amount of attention in the deep learning community due to the development of scalable approximate Bayesian inference techniques. There are several advantages of using a Bayesian approach: parameter and prediction uncertainties become easily available, facilitating more rigorous statistical analysis. Furthermore, prior knowledge can be incorporated. However, the construction of scalable techniques that combine both structural and parameter uncertainty remains a challenge. In this paper, we apply the concept of model uncertainty as a framework for structural learning in BNNs and, hence, make inferences in the joint space of structures/models and parameters. Moreover, we suggest an adaptation of a scalable variational inference approach with reparametrization of marginal inclusion probabilities to incorporate the model space constraints. Experimental results on a range of benchmark datasets show that we obtain comparable accuracy results with the competing models, but based on methods that are much more sparse than ordinary BNNs.
Storvik, Geir Olve; Palomares, Alfonso Diz-Lois; Engebretsen, Solveig; Rø, Gunnar Øyvind Isaksson; Engø-Monsen, Kenth; Kristoffersen, Anja Bråthen; Blasio, Birgitte Freiesleben De og Frigessi, Arnoldo. (2023).
A sequential Monte Carlo approach to estimate a time-varying reproduction number in infectious disease models: the Covid-19 case.
Vis sammendrag
Abstract The Covid-19 pandemic has required most countries to implement complex sequences of non-pharmaceutical interventions, with the aim of controlling the transmission of the virus in the population. To be able to take rapid decisions, a detailed understanding of the current situation is necessary. Estimates of time-varying, instantaneous reproduction numbers represent a way to quantify the viral transmission in real time. They are often defined through a mathematical compartmental model of the epidemic, like a stochastic SEIR model, whose parameters must be estimated from multiple time series of epidemiological data. Because of very high dimensional parameter spaces (partly due to the stochasticity in the spread models) and incomplete and delayed data, inference is very challenging. We propose a state-space formalization of the model and a sequential Monte Carlo approach which allow to estimate a daily-varying reproduction number for the Covid-19 epidemic in Norway with sufficient precision, on the basis of daily hospitalization and positive test incidences. The method was in regular use in Norway during the pandemics and appears to be a powerful instrument for epidemic monitoring and management.
Hubin, Aliaksandr; Storvik, Geir og Frommlet, Florian. (2021).
Flexible Bayesian Nonlinear Model Configuration.
Vis sammendrag
Regression models are used in a wide range of applications providing a powerful scientific tool for researchers from different fields. Linear, or simple parametric, models are often not sufficient to describe complex relationships between input variables and a response. Such relationships can be better described through flexible approaches such as neural networks, but this results in less interpretable models and potential overfitting. Alternatively, specific parametric nonlinear functions can be used, but the specification of such functions is in general complicated. In this paper, we introduce a flexible approach for the construction and selection of highly flexible nonlinear parametric regression models. Nonlinear features are generated hierarchically, similarly to deep learning, but have additional flexibility on the possible types of features to be considered. This flexibility, combined with variable selection, allows us to find a small set of important features and thereby more interpretable models. Within the space of possible functions, a Bayesian approach, introducing priors for functions based on their complexity, is considered. A genetically modified mode jumping Markov chain Monte Carlo algorithm is adopted to perform Bayesian inference and estimate posterior probabilities for model averaging. In various applications, we illustrate how our approach is used to obtain meaningful nonlinear models. Additionally, we compare its predictive performance with several machine learning algorithms.
Storvik, Geir Olve. (2021).
Bayesian Approaches to Neural Networks. UiT SFI Visual Intelligence
NVA
Faglig foredrag
Hubin, Aliaksandr; Storvik, Geir Olve og Frommlet, Florian. (2020).
A novel algorithmic approach to Bayesian Logic Regression. ISBA, IMS
NVA
Vitenskapelig foredrag
Vis sammendrag
Logic regression was developed more than a decade ago as a tool to construct predictors from Boolean combinations of binary covariates. It has been mainly used to model epistatic effects in genetic association studies, which is very appealing due to the intuitive interpretation of logic expressions to describe the interaction between genetic variations. Nevertheless logic regression has (partly due to computational challenges) remained less well known than other approaches to epistatic association mapping. Here we will adapt an advanced evolutionary algorithm called GMJMCMC (Genetically modified Mode Jumping Markov Chain Monte Carlo) to perform Bayesian model selection in the space of logic regression models. After describing the algorithmic details of GMJMCMC we perform a comprehensive simulation study that illustrates its performance given logic regression terms of various complexity. Specifically GMJMCMC is shown to be able to identify three-way and even four-way interactions with relatively large power, a level of complexity which has not been achieved by previous implementations of logic regression. We apply GMJMCMC to reanalyze QTL (quantitative trait locus) mapping data for Recombinant Inbred Lines in Arabidopsis thaliana and from a backcross population in Drosophila where we identify several interesting epistatic effects. The method is implemented in an R package which is available on github.
Hubin, Aliaksandr; Storvik, Geir Olve og Frommlet, Florian. (2020).
Rejoinder for the discussion of the paper "A Novel Algorithmic Approach to Bayesian Logic Regression".
Vis sammendrag
Logic regression was developed more than a decade ago as a tool to construct predictors from Boolean combinations of binary covariates. It has been mainly used to model epistatic effects in genetic association studies, which is very appealing due to the intuitive interpretation of logic expressions to describe the interaction between genetic variations. Nevertheless logic regression has (partly due to computational challenges) remained less well known than other approaches to epistatic association mapping. Here we will adapt an advanced evolutionary algorithm called GMJMCMC (Genetically modified Mode Jumping Markov Chain Monte Carlo) to perform Bayesian model selection in the space of logic regression models. After describing the algorithmic details of GMJMCMC we perform a comprehensive simulation study that illustrates its performance given logic regression terms of various complexity. Specifically GMJMCMC is shown to be able to identify three-way and even four-way interactions with relatively large power, a level of complexity which has not been achieved by previous implementations of logic regression. We apply GMJMCMC to reanalyze QTL (quantitative trait locus) mapping data for Recombinant Inbred Lines in Arabidopsis thaliana and from a backcross population in Drosophila where we identify several interesting epistatic effects. The method is implemented in an R package which is available on github.
Hubin, Aliaksandr; Storvik, Geir Olve; Grini, Paul Eivind og Butenko, Melinka Alonso. (2020).
A Bayesian binomial regression model with latent gaussian processes for modelling DNA methylation.
Vis sammendrag
Epigenetic observations are represented by the total number of reads from a given pool of cells and the number of methylated reads, making it reasonable to model this data by a binomial distribution. There are numerous factors that can influence the probability of success in a particular region. Moreover, there is a strong spatial (alongside the genome) dependence of these probabilities. We incorporate dependence on the covariates and the spatial dependence of the methylation probability for observations from a pool of cells by means of a binomial regression model with a latent Gaussian field and a logit link function. We apply a Bayesian approach including prior specifications on model configurations. We run a mode jumping Markov chain Monte Carlo algorithm (MJMCMC) across different choices of covariates in order to obtain the joint posterior distribution of parameters and models. This also allows finding the best set of covariates to model methylation probability within the genomic region of interest and individual marginal inclusion probabilities of the covariates.
Hubin, Aliaksandr; Storvik, Geir Olve; Grini, Paul Eivind og Butenko, Melinka Alonso. (2019).
Bayesian binomial regression model with a latent Gaussian field for analysis of epigenetic data.
NVA
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
Hubin, Aliaksandr og Storvik, Geir Olve. (2019).
Combining Model and Parameter Uncertainty in Bayesian Neural Networks. Norwegian Open AI Lab and Norwegian University of Science an
NVA
Vitenskapelig foredrag
Vis sammendrag
Bayesian neural networks (BNNs) have recently regained a significant amount of attention in the deep learning community due to the development of scalable approximate Bayesian inference techniques. There are several advantages of using Bayesian approach: Parameter and prediction uncertainty become easily available, facilitating rigid statistical analysis. Furthermore, prior knowledge can be incorporated. However so far there have been no scalable techniques capable of combining both model (structural) and parameter uncertainty. In this paper we introduce the concept of model uncertainty in BNNs and hence make inference in the joint space of models and parameters. Moreover, we suggest an adaptation of a scalable variational inference approach with reparametrization of marginal inclusion probabilities to incorporate the model space constraints. Finally, we show that incorporating model uncertainty via Bayesian model averaging and Bayesian model selection allows to drastically sparsify the structure of BNNs without significant loss of predictive power.
Hubin, Aliaksandr og Storvik, Geir Olve. (2019).
Combining Model and Parameter Uncertainty in Bayesian Neural Networks. ECOSTA (workshop), CRoNoS COST Action (Spring Course and fin
NVA
Faglig foredrag
Storvik, Geir Olve og Hubin, Aliaksandr. (2019).
Combining model and parameter uncertainty in Bayesian neural networks. ERCIM Working Group on Computational and Methodological Stat
NVA
Vitenskapelig foredrag
Hubin, Aliaksandr og Storvik, Geir Olve. (2019).
Combining Model and Parameter Uncertainty in Bayesian Neural Networks. Open Data Science (ODS.AI)
NVA
Faglig foredrag
Vis sammendrag
Bayesian neural networks (BNNs) have recently regained a significant amount of attention in the deep learning community due to the development of scalable approximate Bayesian inference techniques. There are several advantages of using Bayesian approach: Parameter and prediction uncertainty become easily available, facilitating rigid statistical analysis. Furthermore, prior knowledge can be incorporated. However so far there have been no scalable techniques capable of combining both model (structural) and parameter uncertainty. In this paper we introduce the concept of model uncertainty in BNNs and hence make inference in the joint space of models and parameters. Moreover, we suggest an adaptation of a scalable variational inference approach with reparametrization of marginal inclusion probabilities to incorporate the model space constraints. Finally, we show that incorporating model uncertainty via Bayesian model averaging and Bayesian model selection allows to drastically sparsify the structure of BNNs without significant loss of predictive power.
Hubin, Aliaksandr og Storvik, Geir Olve. (2019).
Combining Model and Parameter Uncertainty in Bayesian Neural Networks. Big insight, UiO, NR
NVA
Faglig foredrag
Hubin, Aliaksandr; Storvik, Geir Olve; Grini, Paul Eivind og Butenko, Melinka Alonso. (2019).
Bayesian binomial regression model with a latent Gaussian field for analysis of epigenetic data. Belarusian State University, Vienna University of Technology
NVA
Faglig foredrag
Hubin, Aliaksandr og Storvik, Geir Olve. (2019).
Combining Model and Parameter Uncertainty in Bayesian Neural Networks. Belarusian State University, Vienna University of Technology
NVA
Faglig foredrag
Hubin, Aliaksandr; Storvik, Geir Olve og Frommlet, Florian. (2018).
Deep Bayesian regression models. NORBIS
NVA
poster
Hubin, Aliaksandr; Storvik, Geir Olve og Frommlet, Florian. (2018).
A Novel Algorithmic Approach to Bayesian Logic Regression.
Vis sammendrag
Logic regression was developed more than a decade ago as a tool to
construct predictors from Boolean combinations of binary covariates. It has been
mainly used to model epistatic effects in genetic association studies, which is very
appealing due to the intuitive interpretation of logic expressions to describe the interaction
between genetic variations. Nevertheless logic regression has (partly due
to computational challenges) remained less well known than other approaches to
epistatic association mapping. Here we will adapt an advanced evolutionary algorithm
called GMJMCMC (Genetically modified Mode Jumping Markov Chain
Monte Carlo) to perform Bayesian model selection in the space of logic regression
models. After describing the algorithmic details of GMJMCMC we perform
a comprehensive simulation study that illustrates its performance given logic regression
terms of various complexity. Specifically GMJMCMC is shown to be able
to identify three-way and even four-way interactions with relatively large power,
a level of complexity which has not been achieved by previous implementations
of logic regression. We apply GMJMCMC to reanalyze QTL (quantitative trait
locus) mapping data for Recombinant Inbred Lines in Arabidopsis thaliana and
from a backcross population in Drosophila where we identify several interesting
epistatic effects. The method is implemented in an R package which is available
on github.
Hubin, Aliaksandr; Storvik, Geir Olve og Frommlet, Florian. (2018).
Deep Bayesian regression models. University of Oslo
NVA
Vitenskapelig foredrag
Rognebakke, Hanne; Hirst, David; Aanes, Sondre og Storvik, Geir Olve. (2016).
Catch-at-age - Version 4.0: Technical Report.
Vis sammendrag
The Norwegian Computing Center and the Institute of Marine Research have over years developed a Bayesian hierarchical model to estimate the catch-at-age of fish. Such a model enables us to obtain estimates of the catch-at-age with appropriate uncertainty. This is considered as essential input in most age structured stock assessment processes. Recent improvements of the model include modelling a haulsize effect in the proportionat-age model. The model is implemented in C with an R interface. This note gives a thorough description of the model and the simulation algorithm corresponding to version 4.0 of the program.
Storvik, Geir Olve; Løland, Anders; Lykkja, Ola Martin og Gjevestad, Jon Glenn Omholt. (2016).
SAVE – tracking vehicle movements for toll object detection using particle filter.
NVA
Rapport
Storvik, Geir Olve; Aanes, Sondre og Maisha, Peter Nyangweso. (2015).
Estimation of fish abundance and demography in the Barents sea. International Biometric Society
NVA
Vitenskapelig foredrag
Maisha, Peter Nyangweso; Storvik, Geir Olve og Aanes, Sondre. (2015).
A State-Space Model for Abundance Estimation from Bottom Trawl Data with Applications to Norwegian Winter Survey.
Vis sammendrag
We study a hierarchical dynamic state-space model for abundance estimation. A generic data fusion approach for combining computer simulated posterior samples of catch output data with observed research survey indices using sequential importance sampling is presented. Posterior samples of catch generated from a computer software are used as a primary source of input data through which fisheries dependent information is mediated. Direct total stock abundance estimates are obtained without the need to estimate any intermediate parameters such as catchability and mortality. Numerical results of a simulation study show that our method provides a useful alternative to existing methods. We apply the method to data from the Barents Sea Winter survey for Northeast Arctic cod (Gadus morhua). The results based on our method are comparable to results based on current methods.
Marques, Reinaldo A. Gomes og Storvik, Geir Olve. (2014).
Reweighting Schemes Based on Particle Methods.
Vis sammendrag
Sequential Monte Carlo methods are widely used to deal with the intractability of complex models including state space models. Their aim is to approximate the distribution of interest by a set of properly weighted samples. To control the weight degeneracy, the resample step has been proposed as an inexpensive alternative to avoid the collapse of particle filter algorithms. When the sample becomes too poor with successive use of resample steps, MCMC moves have been added in particle filter algorithms in order to make the identical samples diverge. In this work we consider strategies where we first perform a moves step, and then we update the weights for reweighting the particles. The validity of this approach is based on the commonly used trick of working on an artificial extended distribution having the target distribution as marginal combined with the use of backwards kernels. By updating the weights via a diversification step, this approach can make their empirical distribution less skewed increasing the effective sample size.
Hirst, David; Storvik, Geir Olve; Rognebakke, Hanne; Aldrin, Magne; Aanes, Sondre og Vølstad, Jon Helge. (2012).
A Bayesian modelling framework for the estimation of catch-at-age of commercially harvested fish species.
Aldrin, Magne; Mortensen, Bjørnar Tumanjan; Storvik, Geir Olve; Nedreaas, Kjell; Aglen, Asgeir og Aanes, Sondre. (2012).
Improving management decisions by predicting fish bycatch in the Barents Sea shrimp fishery.
Vis sammendrag
When the bycatch of juvenile fish within the Barents Sea shrimp fishery is too large, the area is closed to fishing for a certain period. Bycatch is estimated from sampled trawl hauls, for which the shrimp yield is recorded, along with the total number of various bycatch fish species. At present, bycatch estimation is based on a simple estimator, the sum of the number of fish caught within the area of interest within a small time window, divided by the corresponding shrimp yield (in weight). No historical data are used. A model-based estimation is proposed in which spatio-temporal models are constructed for the variation in both the yield of shrimp and the amount of bycatch in space and time. The main effects are described through generalized additive models, and local dependence structures are specified through correlated random effects. Model estimation includes historical and recent data. Experiments with both simulated and real data show that the model-based estimator outperforms the present simple estimator when a low or moderate number of samples (e. g. <20) is available, whereas the two estimators are equally good when the number of samples is high.
Rognebakke, Hanne; Hirst, David; Storvik, Geir og Aldrin, Magne. (2011).
Catch-at-age for multiple stocks: Modelling Skrei and Coastal Cod simultaneously.
Vis sammendrag
The Norwegian Computing Center and the Institute of Marine Research have over years developed a Bayesian hierarchical model to estimate the catch-at-age of fish for a single stock. This has been extended to model multiple stocks, like Atlantic Cod and Coastal Cod. This note presents the multiple stock model, which includes both age reading error and classification error.
Rognebakke, Hanne Therese Wist; Hirst, David; Aldrin, Magne og Storvik, Geir. (2011).
Modelling catch at age for multiple stock. Norsk statistisk forening
NVA
Vitenskapelig foredrag
Rognebakke, Hanne Therese Wist; Hirst, David og Storvik, Geir Olve. (2011).
Catch-at-age - Version 2.0: Technical Report.
Vis sammendrag
The Norwegian Computing Center and the Institute of Marine Research have over years developed a Bayesian hierarchical model to estimate the catch-at-age of fish. The model is implemented in C with an R interface. This note gives a thorough description of the model and the simulation algorithm corresponding to version 2.0 of the program.
Villar, Jaime Otero; Jensen, Arne Johan; L'Abée-Lund, Jan Henning; Stenseth, Nils Christian; Storvik, Geir Olve og Vøllestad, Leif Asbjørn. (2010).
Contemporary ocean warming and freshwater conditions contribute to delay the completion of maturation in Atlantic salmon throughout the Norwegian range of distribution. ICES
NVA
Vitenskapelig foredrag
Villar, Jaime Otero; Antonsson, Thorulfur; Armstrong, John; Arnason, Fridthjofur; Arnekleiv, Jo Vegar; Baglinière, Jean-Luc; Caballero, Pablo; Castro-Santos, Ted; Dempson, Brian; Erkinaro, Jaakko; Gudjonsson, Sigurdur; Hvidsten, Nils Arne; Jensen, Arne Johan; Jokikokko, Erkki; Jonsson, Ingi Runar; Kocik, John; L'Abée-Lund, Jan Henning; Lamberg, Anders; Letcher, Benjamin; Niemela, Eero; Romakkaniemi, Atso; Russell, Ian; Stenseth, Nils Christian; Storvik, Geir Olve; Veselov, Alexey og Vøllestad, Leif Asbjørn. (2010).
Environmental effects on ocean entry of Atlantic salmon (Salmo salar) smolt across its range of distribution. ICES
NVA
poster
Steinbakk, Gunnhildur Högnadóttir og Storvik, Geir Olve. (2009).
Posterior Predictive p-values in Bayesian Hierarchical Models.
Storvik, Bård; Storvik, Geir Olve og Fjørtoft, Roger. (2009).
On the Combination of Multisensor Data Using Meta-Gaussian Distributions.
Hirst, David; Storvik, Geir Olve; Rognebakke, Hanne og Aldrin, Magne. (2009).
A Bayesian hierarchical modelling approach to estimating landings- and discards-at-age. ICES
NVA
Vitenskapelig foredrag
Storvik, Bård; Storvik, Geir og Fjørtoft, Roger Sverre. (2008).
On the combination of correlated images using meta-Gaussian distributions.
NVA
Vitenskapelig artikkel
Hirst, David; Rognebakke, Hanne Therese Wist; Storvik, Geir Olve og Aldrin, Magne. (2007).
Implementing the catch-at-age program for FRS Aberdeen.
NVA
Rapport
Storvik, Geir Olve og Egeland, Thore. (2007).
The DNA database search controversy revisited: bridging the Bayesian-Frequentist gap.
Hjermann, Dag Øystein; Melsom, Arne; Dingsør, Gjert Endre; Durant, Joël M.; Eikeset, Anne Maria; Røed, Lars Petter; Ottersen, Geir; Storvik, Geir og Stenseth, Nils Christian. (2007).
Fish and oil in the Lofoten-Barents Sea system: synoptic review of the effect of oil spills on fish populations.
NVA
Vitenskapelig artikkel
Storvik, Geir Olve. (2005).
Modell-evaluering i Bayesianske modeller. Norsk Kjemisk Selskap
NVA
Vitenskapelig foredrag
Hirst, David; Storvik, Geir Olve; Aldrin, Magne og Aanes, Sondre. (2005).
Bayesian estimation of catc-at-age using data from several sources.
NVA
Rapport
Storvik, Geir Olve. (2005).
Evaluation of evolutionary models in phylogenetic inferences. The bioportal project
NVA
Vitenskapelig foredrag
Hirst, David; Storvik, Geir Olve; Aldrin, Magne Tommy; Aanes, Sondre og Huseby, Ragnar Bang. (2005).
Estimating catch-at-age by combining data from different sources.
Vis sammendrag
Estimating the catch-at-age of commercial fish species is an important part of the quota-setting process for many different species and almost all countries with a fishing fleet. Current procedures are usually very time-consuming and somewhat ad hoc, and the estimates have no measure of uncertainty. We previously developed a method for catch-at-age of Norwegian Atlantic cod (Gadus morhua), but this only considered aged fish sampled randomly from random hauls. In most countries, the sampling scheme is not so simple. There are usually a very large number of length-only samples from which the age must be estimated using an age-length relationship, and often some or all of the age samples are collected from data that are first stratified by length. This adds considerably to the difficulties in the estimation. In this paper, we model the three different kinds of data simultaneously using a development of our earlier Bayesian hierarchical model. This enables us to obtain estimates of the catch-at-age with appropriate uncertainty and also to provide advice on how best to sample data in the future. The data types are random samples of age, length, and weight; age and weight stratified by length; and length only.
Storvik, Geir Olve; Fjørtoft, Roger Sverre og Solberg, Anne H S. (2005).
A Bayesian approach to classification of multi-scale remote sensing data.
Vis sammendrag
Several earth observation satellites acquire image bands with different spatial resolutions, e.g., a panchromatic band with high resolution and spectral bands with lower resolution. Likewise, we often face the problem of different resolutions when performing joint analysis of images acquired by different satellites. This work presents models and methods for classification of multiresolution images. The approach is based on the concept of a reference resolution, corresponding to the highest resolution in the dataset. Prior knowledge about the spatial characteristics of the classes is specified through a Markov random field model at the reference resolution. Data at coarser scales are modeled as mixed pixels by relating the observations to the classes at the reference resolution. A Bayesian framework for classification based on this multiscale model is proposed. The classification is realized by an iterative conditional modes (ICM) algorithm. The parameter estimation can be based both on a training set and on pixels with unknown class. A computationally efficient scheme based on a combination of the ICM and the expectation-maximization algorithm is proposed. Results obtained on simulated and real satellite images are presented.
Storvik, Geir. (2005).
Hvorfor Data + Data = Sant - Fra telling av fisk til DNA-profiler av voldsmenn.
NVA
Faglig foredrag
Storvik, Geir Olve. (2005).
Bayesian assessment of Bowhead whales. CEES, UiO
NVA
Vitenskapelig foredrag
Storvik, Geir Olve og Egeland, Thore. (2005).
Identification of DNA profiles from mixture samples. International Biometric Society
NVA
Vitenskapelig foredrag
Hirst, David; Aanes, S; Storvik, Geir Olve; Huseby, Ragnar Bang og Tvete, Ingunn. (2004).
Estimating catch at age from market sampling data by using a Bayesian hierarchical model.
Vis sammendrag
The paper develops a Bayesian hierarchical model for estimating the catch at age of cod landed in Norway. The model includes covariate effects such as season and gear, and can also account for the within-boat correlation. The hierarchical structure allows us to account properly for the uncertainty in the estimates.
Aldrin, Magne; Huseby, Ragnar Bang; Høst, Gudmund; Løland, Anders og Storvik, Geir Olve. (2004).
Spatial-Temporal Uncertainty for Survey-Based Abundance Estimates.
NVA
Rapport
Storvik, Bård; Storvik, Geir og Fjørtoft, Roger. (2003).
Joint distributions for correlated radar images.
NVA
Vitenskapelig artikkel
Hirst, David og Storvik, Geir. (2003).
Estimating Critical load exceedance by combining the EMEP model with data from measurement stations.
NVA
Vitenskapelig artikkel
Storvik, Bård; Storvik, Geir og Fjørtoft, Roger Sverre. (2003).
Joint distribution of correlated radar images.
NVA
Rapport
Storvik, Bård; Storvik, Geir Olve og Fjørtoft, Roger Sverre. (2003).
Joint distribution for correlated radar images.
NVA
Vitenskapelig foredrag
Storvik, Bård; Storvik, Geir Olve og Fjørtoft, Roger. (2003).
Joint distributions for multi-temporal series of radar images.
NVA
Vitenskapelig foredrag
Storvik, Geir Olve; Fjørtoft, Roger Sverre og Solberg, Anne H S. (2003).
Parameter estimation and classification of multi-scale remote sensing data.
NVA
Vitenskapelig foredrag
Hirst, David; Storvik, Geir Olve; Syversveen, Anne Randi og Syversveen, Anne Randi. (2003).
A hierarchical modelling approach to combining environmental data at different resolutions.
NVA
Vitenskapelig artikkel
Hirst, David; Storvik, Geir og Syversveen, Anne Randi. (2003).
A hierarchical modelling approach to combining environmental data at different scales.
NVA
Vitenskapelig foredrag
Solberg, Anne H S; Storvik, Geir Olve og Fjørtoft, Roger Sverre. (2002).
A Comparison of Criteria for Decision Fusion and Parameter Estimation in Statistical Multisensor Image Classification.
NVA
Vitenskapelig foredrag
Dahl, Geir; Storvik, Geir Olve; Fadnes, Alice; Dahl, Gustav og Fadnes, A.. (2002).
Large-scale integer programs in image analysis.
NVA
Vitenskapelig artikkel
Storvik, Geir; Frigessi, Arnoldo og Hirst, David. (2002).
Space-time Gaussian fields and their time-autoregressive representation.
NVA
Vitenskapelig artikkel
Storvik, Geir Olve; Fjørtoft, Roger og Solberg, Anne H S. (2002).
Integration of multisource image data at different resolutions and time points. A mathematical framework for EOtools.
NVA
Rapport
Fjørtoft, Roger Sverre; Solberg, Anne H S og Storvik, Geir Olve. (2002).
A new approach to statistical multisensor image classification.
NVA
Vitenskapelig foredrag
Storvik, Geir Olve; Frigessi, Arnoldo og Hirst, David. (2001).
Stationary space time Gaussian fields and their time autoregressive representation.
Storvik, Geir Olve. (2001).
Particle filters for state space models with the presence of unknown static parameters.
Hirst, David; Tvete, Ingunn; Storvik, Geir Olve og Aanes, Sondre. (2001).
Estimating catch-at-age from market sampling data using a Bayesian hierarchical model.
NVA
Vitenskapelig foredrag
Bølviken, Erik og Storvik, Geir Olve. (2001).
Deterministic and stochastic particle filters in state space models.
NVA
Vitenskapelig Kapittel/Artikkel/Konferanseartikkel
Storvik, Geir. (2000).
Some further topics on Monte Carlo methods for dynamic Bayesian problems.
NVA
Vitenskapelig foredrag
Storvik, Geir og Dahl, Geir. (2000).
Lagrangian based methods for finding {MAP} solutions for {MRF} models.
NVA
Vitenskapelig artikkel
Storvik, Geir Olve. (2000).
Structural Modeling for Spatial and Spatio-temporal processes.
NVA
Vitenskapelig foredrag
Huseby, Ragnar Bang; Storvik, Geir Olve og Huseby, Ragnar Bang. (2000).
Recognition of pipes using cross-profile information: A feasibility study.
NVA
Rapport
Solberg, Anne H S; Storvik, Geir Olve; Solberg, Rune og Volden, Espen. (1999).
Automatic detection of oil spills in ERS SAR images.
Vis sammendrag
We present algorithms for the automatic detection of oil spills in SAR images. The developed framework consists of first detecting dark spots in the
image, then computing a set of features for each dark spot, before the spot is classified as either an oil slick or a ''lookalike'' (other oceanographic
phenomena which resemble oil slicks), The classification rule is constructed by combining statistical modeling with a rule-based approach. Prior
knowledge about the higher probability for the presence of oil slicks around ships and oil platforms is incorporated into the model, In addition, knowledge
about the external conditions like mind level and slick surroundings are taken into account. The presented algorithms are tested on 84 SAR images. The
algorithm can discriminate between oil slicks and lookalikes with high accuracy. 94% of the oil slicks and 99% of the lookalikes mere correctly classified.
Solberg, Anne H. Schistad; Holden, Marit; Storvik, Geir; Teigland, André og Møller-Holst, Johan. (1998).
Seismic image analysis: 3D texture attributes and classification - part III.
NVA
Rapport
Solberg, Anne H S; Storvik, Geir Olve; Teigland, André; Møller-Holst, Johan; Berge, Anker M. og Abrahamsen, Petter. (1997).
Seismic image analysis: A pilot study for investigating the relevance of using image analysis techniques on seismic data.
NVA
Rapport
Tjetland, Bjørn Gunnar; Solberg, Rune; Teigland, André; Storvik, Geir; Thune, Mari; Koren, Hans og Opgård, Helge. (1997).
Formation Evaluation Workstation, FEW, Stage 4, Final report on research and development 1996.
NVA
Rapport
Solberg, Anne H. Schistad; Volden, Espen og Storvik, Geir. (1997).
Slick signature screening: Algorithm improvement.
NVA
Rapport
Holden, Marit; Solberg, Anne H. Schistad; Storvik, Geir; Teigland, André; Møller-Holst, Johan og Berge, Anker M.. (1997).
Seismic image analysis: 3D texture attributes and classification - part II.
NVA
Rapport
Storvik, Geir Olve. (1996).
Bayesian Surface Reconstruction from Noisy Images.
NVA
Vitenskapelig foredrag
Vis sammendrag
Reconstruction of surfaces from images alone is usually difficult due
to noise. Prior information such as smoothness, shape, size etc. may
however be available. The Bayesian framework makes it possible for a
formal integration of such prior information and the observed data.
Furthermore, the development of the Markov Chain Monte Carlo method
makes it possible for simulation from the posterior of the surface
given the observed images.
In Storvik (1994) this approach was applied to reconstruction of
contours of objects in two-dimensional images. Similar approaches has
been used in Mardia and Qian among others. Extending these approaches
to three dimensions and/or time is in principle simple, but both the
cost of implementation and the computer power needed for performing
simulations are high.
In this talk we will discuss the use of Bayesian modeling and
stochastic sampling for reconstruction of surfaces both in two and
three dimensions. Construction of appropriate models, algorithms and
estimation procedures for the parameters involved will be considered.
Examples from medical imaging will be presented.
Mostad, Petter F.; Egeland, Thore; Hjort, Nils Lid og Storvik, Geir. (1996).
Uncertainties of reservoir parameters in a Bayesian framework.
NVA
Rapport
Bølviken, Erik; Storvik, Geir; Høgåsen, Gutorm Thomas og Larsson, Pål Gunnar. (1995).
Textural segmentation in 1D through sequences of double stochastic segments.
NVA
Rapport
Solberg, Anne H S; Bosnes, Vidar; Storvik, Geir Olve og Bosnes, Vidar. (1995).
Tissue classification in MR images based on a mixed pixel model.
NVA
Vitenskapelig foredrag
Storvik, Geir; Aas, Kjersti; Teigland, André og Larsson, Pål. (1995).
Recognition of spikes in EEG-signals: Further experimentations with the Hidden Markov Chain model.
NVA
Rapport
Storvik, Geir Olve. (1994).
A Bayesian approach to dynamic contours through stochastic sampling and simulated annealing.
NVA
Vitenskapelig artikkel
Vis sammendrag
In many applications of image analysis, simply connected objects are to
be located in noisy images. During the last 5-6 years active contour
models have become popular for finding the contours of such objects.
Connected to these models are iterative algorithms for finding the
minimizing energy curves making the curves behave dynamically through
the iterations. These approaches do however have several disadvantages.
The numerical algorithms that are in use constraint the models that can
be used. Furthermore, in many cases only local minima can be achieved.
In this paper, we discuss a method for curve detection based on a fully
Bayesian approach. A model for image contours which allows the number
of nodes on the contours to vary is introduced. Iterative algorithms
based on stochastic sampling is constructed, which make it possible to
simulate samples from the posterior distribution, making estimates and
uncertainty measures of specific quantities available. Further,
simulated annealing schemes making the curve move dynamically towards
the global minimum energy configuration are presented. In theory, no
restrictions on the models are made. In practice, however,
computational aspects must be taken into consideration when choosing
the models. Much more general models than the one used for active
contours may however be applied. The approach is applied to ultrasound
images of the left ventricle and to Magnetic Resonance images of the
human brain, and show promising results.
Storvik, Geir Olve. (1994).
Recognition of spikes in EEG patterns using hidden Markov chains.
NVA
Vitenskapelig foredrag
Storvik, Geir Olve; Bølviken, Erik og Solberg, Anne H S. (1994).
Hidden Markov Chains for recognition of structures in time series.
NVA
Rapport
Solberg, Anne H. Schistad; Bosnes, Vidar og Storvik, Geir. (1994).
Tissue classification in MR images based on a mixed pixel model.
NVA
Vitenskapelig artikkel
Holden, Marit; Aas, Kjersti; Lundervold, Arvid; Milvang, Otto; Storvik, Geir; Sebastini, Giovanni; Godtlibsen, Fred og Kristoffersen, Doris Tove. (1993).
Assessment of tissue perfusion and characterization of lesions in dynamical MRI using contrast agents and statistical analysis - methodological considerations and clinical applications.
NVA
Rapport
Storvik, Geir. (1993).
Statistical analysis of multispectral MR images from the brain.
NVA
Vitenskapelig foredrag
Storvik, Geir og Lundervold, Arvid. (1993).
Segmentation of brain parenchyma and CSF in multispectral MR images of the head.
NVA
Vitenskapelig artikkel
Storvik, Geir og Lundervold, Arvid. (1993).
Thermography in combination with image analysis and pattern recognition techniques as a diagnostic tool after whiplash accident/injury - A feasibility study based on a small sample of patients and healthy volunteers.
NVA
Rapport